MongoDB Blog
Announcements, updates, news, and more
MongoDB AI Course in Partnership with Andrew Ng and DeepLearning.AI
MongoDB is committed to empowering developers and meeting them where they are. With a thriving community of 7 million developers across 117 regions, MongoDB has become a cornerstone in the world of database technology. Building on this foundation, we're excited to announce our collaboration with AI pioneer Andrew Ng and DeepLearning.AI, a leading educational technology company specializing in AI and machine learning. Together, we've created an informative course that bridges the gap between database technology and modern AI applications, further enhancing our mission to support developers in their journey to build innovative solutions. Introducing "Prompt Compression and Query Optimization" MongoDB’s latest course on DeepLearning.AI, Prompt Compression and Query Optimization , covers the prominent form factor of modern AI applications today: Retrieval Augmented Generation (RAG) . This course showcases how MongoDB Atlas Vector Search capabilities enable developers to build sophisticated AI applications, leveraging MongoDB as an operational and vector database. To ensure that learners taking this course are not just introduced to vector search, the course presents an approach to reducing the operational cost of running AI applications in production by a technique known as prompt compression. “RAG, or retrieval augmented generation, has moved from being an interesting new idea a few months ago to becoming a mainstream large-scale application.” — Andrew Ng, DeepLearning.AI Key course highlights RAG Applications: Learn to build and optimize the most prominent form of AI applications using MongoDB Atlas and the MongoDB Query Language(MQL). MongoDB Atlas Vector Search: Leverage the power of vector search for efficient information retrieval. MongoDB Document Model: Explore MongoDB's flexible, JSON-like document model, which represents complex data structures and is ideal for storing and querying diverse AI-related data. Prompt Compression: Use techniques to reduce the operational costs of AI applications in production environments. In this course, you'll learn techniques to enhance your RAG applications' efficiency, search relevance, and cost-effectiveness. As AI applications become more sophisticated, efficient data retrieval and processing becomes crucial. This course bridges the gap between traditional database operations and modern vector search capabilities, enabling you to confidently build robust, scalable AI applications that can handle real-world challenges. MongoDB's document model: The perfect fit for AI A key aspect of this course is that it introduces learners to MongoDB's document model and its numerous benefits for AI applications: Python-Compatible Structure: MongoDB's BSON format aligns seamlessly with Python dictionaries, enabling effortless data representation and manipulation. Schema Flexibility: Adapt to varied data structures without predefined schemas, matching the dynamic nature of AI applications. Nested Data Structures: Easily represent complex, hierarchical data often found in AI models and datasets. Efficient Data Ingestion: Directly ingest data without complex transformations, speeding up the data preparation process. Leveraging the combined insights from MongoDB and DeepLearning.AI, this course offers a perfect blend of practical database knowledge and advanced AI concepts. Who should enroll? This course is ideal for developers who: Are familiar with vector search concepts Building RAG applications and Agentic Systems Have a basic understanding of Python and MongoDB and are curious about AI Want to optimize their RAG applications for better performance and cost-efficiency This course offers an opportunity to grasp techniques in AI application development. You'll gain the skills to build more efficient, powerful, cost-effective RAG applications, from advanced query optimization to innovative prompt compression. With hands-on code, detailed walkthroughs, and real-world applications, you'll be equipped to tackle complex AI challenges using MongoDB's robust features. Take advantage of this chance to stay ahead in the rapidly evolving field of AI. Whether you're a seasoned developer or just starting your AI journey, this course will provide invaluable insights and practical skills to enhance your capabilities. Improve your AI application development skills with MongoDB's practical course. Learn to build efficient RAG applications using vector search and prompt compression. Enroll now and enhance your developer toolkit.
Building Gen AI with MongoDB & AI Partners | July 2024
My colleague Richmond Alake recently published an article about the evolution of the AI stack that breaks down the “comprehensive collection of integrated tools, solutions, and components designed to streamline the development and management of AI applications.” It’s a good read, and Richmond—who’s an AI/ML expert and developer advocate—explains clearly how the modern AI stack evolved from a set of disparate tools to the (beautifully) interdependent ecosystem on which AI development relies today. “The modern AI stack represents an evolution from the fragmented tooling landscape of traditional machine learning to a more cohesive and specialized ecosystem optimized for the era of LLMs and gen AI,” Richmond writes. In other words, this cohesive ecosystem is aimed at ensuring end-to-end interoperability and seamless developer experiences, both of which are of utmost importance when it comes to AI innovation (and software innovation overall). Empowering developer innovation is exactly what MongoDB is all about—from streamlining how developers build modern applications, to the blog post you’re reading now, to the news that the MongoDB AI Applications Program (MAAP) is now generally available. In particular, the MAAP ecosystem represents leaders from every part of the AI stack who will provide customer service and support, and who will work with them to ensure smooth integrations—with the ultimate aim of helping them build gen AI applications with confidence. As the saying goes, it takes a village. Welcoming new AI partners Because the AI ecosystem is constantly evolving, we're always working to ensure that customers can seamlessly integrate with the latest cohort of industry-leading companies. In July we welcomed nine new AI partners that offer product integrations with MongoDB. Read on to learn more about each great new partner! Enkrypt AI Enkrypt AI secures enterprises against generative AI risks with its comprehensive security platform that detects threats, removes vulnerabilities, and monitors performance for continuous insights. The solution enables organizations to accelerate AI adoption while managing risk and minimizing brand damage. Sahil Agarwal, CEO of Enkrypt AI said, “We are thrilled to announce our strategic partnership with MongoDB, to help companies secure their RAG workflows for faster production deployment. Together, Enkrypt AI and MongoDB are dedicated to delivering unparalleled safety and performance, ensuring that companies can leverage AI technologies with confidence and improved trust.” FriendliAI FriendliAI’s mission is to empower organizations to harness the full potential of their generative AI models with ease and cost efficiency. By eliminating the complexities of generative AI serving, FriendliAI aims to empower more companies to achieve innovation with generative AI. “We’re excited to partner with MongoDB to empower companies in testing and optimizing their RAG features for faster production deployment,” said Byung-Gon Chon, CEO and co-founder of FriendliAI. “MongoDB simplifies the launch of a scalable vector database with operational data. Our collaboration streamlines the entire RAG development lifecycle, accelerating time to market and enabling companies to deliver real value to their customers more swiftly.” HoneyHive HoneyHive helps organizations continuously debug, evaluate, and monitor AI applications, and ship new AI features faster and with confidence. "We’re thrilled to announce our partnership with MongoDB, which addresses a critical challenge in GenAI deployment—the gap between prototyping and production-ready RAG systems,” said Mohak Sharma, CEO of HoneyHive. “By integrating HoneyHive's evaluation and monitoring capabilities with MongoDB's robust vector database, we're enabling developers to build, test, and deploy RAG applications with greater confidence. This collaboration provides the necessary tools for continuous quality assurance, from development through to production. For companies aiming to leverage gen AI responsibly and at scale, our combined solution offers a pragmatic path to faster, more reliable deployment." Iguazio The Iguazio AI platform operationalizes and de-risks ML & gen AI applications at scale so organizations can implement AI effectively and responsibly in live business environments. “We're delighted to expand our partnership with MongoDB into the gen AI domain, jointly helping enterprises build, deploy and manage gen AI applications in live business environments with our gen AI Factory,” said Asaf Somekh, co-founder and CEO of Iguazio (acquired by McKinsey). “Together, we mitigate the challenges of scaling gen AI and minimizing risk with built-in guardrails. Our seamlessly integrated technologies enable enterprises to realize the potential of gen AI and turn their AI strategy into real business impact." Netlify Netlify is the essential platform for the delivery of exceptional and dynamic web experiences, without limitations. The Netlify Composable Web Platform simplifies content orchestration, streamlines and unifies developer workflow, and enables website speed and agility for enterprise teams. "Netlify is excited to join forces with MongoDB to help companies test and optimize their RAG features for faster production deployment,” said Dana Lawson, Chief Technical Officer at Netlify. “MongoDB has made it easy to launch a scalable vector database with operational data, while Netlify enhances the deployment process and speed to production. Our collaboration streamlines the development lifecycle of RAG applications, decreasing time to market and helping companies deliver real value to customers faster." Render Render helps software teams ship products fast and at any scale. The company hosts applications for customers that range from solopreneurs, small agencies, and early stage startups, to mature, scaling businesses with services deployed around the world, all with a relentless commitment to reliability and uptime. Jess Lin, Developer Advocate at Render, said, “We’re thrilled to join forces with MongoDB to help companies effortlessly deploy and scale their applications—from their first user to their billionth. Render and MongoDB Atlas both empower engineers to focus on developing their products, not their infrastructure. Together, we're streamlining how engineers build full-stack apps, which notably include new AI applications that use RAG.” Superlinked Superlinked is a compute framework that helps MongoDB Atlas Vector Search work at the level of documents, rather than individual properties, enabling MongoDB customers to build high-quality RAG, Search, and Recommender systems with ease. “We're thrilled to join forces with MongoDB to help companies build vector search solutions for complex datasets,” said Daniel Svonava, CEO of Superlinked. “MongoDB makes it simple to manage operational data and a scalable vector index in one place. Our collaboration brings the operational data into the vector embeddings themselves, making the joint system able to answer multi-faceted queries like “largest clients with exposure to manufacturing risk” and operate the full vector search development cycle, speeding up time to market and helping companies get real value to customers faster." Twelve Labs Twelve Labs builds AI that perceives the world the way humans do. The company models the world by shipping next-generation multimodal foundation models that push the boundaries in video understanding. "We are excited to partner with MongoDB to enable developers and enterprises to build advanced multimodal video understanding applications,” said Jae Lee, CEO of Twelve Labs. “Developers can store Twelve Labs' state-of-the-art video embeddings in MongoDB Atlas Vector Search for efficient semantic video retrieval—which enables video recommendations, data curation, RAG workflows, and more. Our collaboration supports native video processing and ensures high-performance & low latency for large-scale video datasets." Upstage Upstage specializes in delivering above-human-grade performance AI solutions for enterprises, focusing on superior usability, customizability, and data privacy. “We are thrilled to partner with MongoDB to provide our enterprise customers with a powerful full-stack LLM solution featuring RAG capabilities,” said Sung Kim, CEO and co-founder of Upstage. “By combining Upstage AI's Document AI, Solar LLM, and embedding models with the robust vector database MongoDB Atlas, developers can create a powerful end-to-end RAG application that's grounded with the enterprise's unstructured data. This application achieves a fast time to value with productivity gains while minimizing the risk of hallucination.” But wait, there's more! To learn more about building AI-powered apps with MongoDB, check out our AI Resources Hub , and stop by our Partner Ecosystem Catalog to read about our integrations with MongoDB’s ever-evolving AI partner ecosystem.
How MongoDB Scales CoPilot AI’s Humanized Sales Interactions
In a world where sales and marketing are the engines behind many tech companies’ growth in a highly competitive landscape, it’s more important than ever that those functions find better and fresher ways to implement personalization into campaigns, sales pitches, and everything in between to reach more customers. CoPilot AI has been at the helm of helping businesses do just that through their AI-powered sales enablement tool, automating personalized interactions to achieve revenue growth, all with the help of MongoDB . CoPilot AI is a software company that helps businesses leverage AI to personalize and automate sales outreach. “We’re looking to humanize digital interactions in a scalable way. That’s our mission, our ethos behind our entire business, which can sometimes seem counterintuitive to people when you think of ‘AI’ and ‘humanize’,” said Scott Morgan, Head of Product Marketing at CoPilot AI. They integrate with platforms that have a high-quality lead base or verified accounts to identify qualified leads and facilitate communication through features like smart replies and sentiment analysis. Today, they predominantly work with LinkedIn as a channel, tapping into the 1B+ professionals globally who interact and conduct business online. “We envision ourselves as having an AI suite of assisting tools that allow business professionals and companies to support their entire sales journey with AI tooling,” said Morgan. CoPilot AI uses five AI pieces (sentiment analysis, reply prediction, smart reply, Meetings Booked AI, and personalized insights) to qualify leads within its lead management platform. Reply prediction gives predictions on which leads are most likely to book meetings with you before you connect with them, while sentiment analysis analyzes replies from leads to determine if they’re interested in continuing the conversation. These features prioritize high-quality leads, boosting success rates. Check out our AI resource page to learn more about building AI-powered apps with MongoDB. The challenge: Scaling AI-powered sales interactions CoPilot AI leverages Random Forest and Chat GPT-3.5 turbo for production and Chat GPT-4 for producing labels and creating models. Their tech stack also includes AWS SageMaker and Azure. However, efficiently managing the data behind these interactions was crucial for scaling their platform. They needed a powerful database to hold and manage their massive system records. Also, they needed a scalable and cost-effective platform that could accommodate their data needs for their growing user bases. Enter, MongoDB Atlas. For most of its 10-year journey, CoPilot AI has been using MongoDB. When evaluating alternative cloud database solutions, CoPilot AI explored Microsoft’s Azure Cosmos DB. While Azure Cosmos DB offered a compelling feature set, its pricing structure didn’t align with CoPilot AI’s specific data access patterns, resulting in high costs. This led them to MongoDB for optimal cost-efficiency for their workload. Building a scalable foundation with MongoDB Atlas CoPilot AI has been using MongoDB since 2013 and started using MongoDB Atlas in 2020. “Everything! System of record, campaigns, message sequences, sending messages, all of that is in MongoDB,” says Takahiro Kato, Principal Engineer at CoPilot AI. CoPilot AI also uses Atlas Data Federation to access its customer information, leads, and campaign conversations. They set up data lake pipelines that go into Data Federation, where their ML engineers pull the data from. They also use Online Archive quite extensively. As a fast-growing startup, CoPilot AI was also able to take advantage of the MongoDB for Startups program , giving them access to free credits and expert technical advice to optimize their usage. “Access to the consultant was quite useful as well. We received advice on how to improve query efficiency, something we’ve been struggling with for a while. In the past, the cost was quite high, our queries were inefficient. As we were going through and fixing those issues, the advice helped,” says Kato. MongoDB empowered CoPilot AI with streamlined development through an intuitive driver and data flexibility via its schema-less design, enabling developers to focus on core functionalities while effortlessly adapting the data model for business growth. CoPilot AI continues to use MongoDB Atlas for multiple reasons, some of which include: Speed and Performance: MongoDB's fast read/write capabilities ensure smooth operation for CoPilot AI's data-intensive operations. Developer Productivity: The C# driver with LINQ support simplifies data access for CoPilot AI's .NET backend, boosting development efficiency. Scalability: MongoDB's flexible schema easily accommodates CoPilot AI's evolving data needs as its user base grows. Cost Optimization: Compared to alternatives, MongoDB offered a more cost-effective option for CoPilot AI's data storage needs. Plus, the MongoDB for Atlas Startups Program provided valuable credits and expert guidance to optimize queries and reduce costs. Key takeaways for developers and businesses MongoDB Atlas securely unifies operational, unstructured, and AI-related data to streamline building AI-enriched applications. Considering leveraging AI in your business? Look no further than MongoDB as your database management solution. If you want to learn more about how you can get started with your next AI project or take your AI capabilities to the next level, you can check out our MongoDB for Artificial Intelligence resources page for the latest best practices that get you started in turning your idea into an AI-driven reality. Also, stop by our Partner Ecosystem Catalog to read about our integrations with MongoDB’s ever-evolving AI partner ecosystem. Head over to our quick-start guide to get started with Atlas Vector Search today. Add vector search to your arsenal for more accurate and cost-efficient RAG applications by enrolling in the MongoDB and DeepLearning.AI course " Prompt Compression and Query Optimization " for free today.
Atlas Stream Processing Adds AWS Regions, VPC Peering, & More!
Since announcing the general availability of Atlas Stream Processing —MongoDB’s native stream processing solution—it’s been exciting to see development teams across technology, retail, and manufacturing begin to run production stream processing workloads critical to their businesses. Today, we're announcing four key updates to Atlas Stream Processing. Support for AWS Regions across the US, Europe, and APAC First, we're thrilled to announce that Atlas Stream Processing now supports eight new AWS regions . This expansion enhances deployment flexibility across the US, Europe, and APAC. Adding these new AWS regions broadens our reach and opens up a world of possibilities for users. We're committed to further expanding our reach by adding more regions and cloud providers in the future. Newly supported regions launched today include: Region AWS Region Name Oregon, USA us-west-2 Sao Paulo, Brazil sa-east-1 Ireland eu-west-1 London, England eu-west-2 Frankfurt, Germany eu-central-1 Mumbai, India ap-south-1 Singapore ap-southeast-1 Sydney, Australia ap-southeast-2 Adding these new AWS regions for Atlas Stream Processing is the latest example of the close partnership between MongoDB and AWS. For example, over the past year, MongoDB announced integrations with Amazon Bedrock and Amazon Q Developer; MongoDB was named an AWS Generative AI Competency Partner ; we launched the MongoDB AI Applications Program —which helps customers rapidly build AI applications—with AWS and other tech leaders; and MongoDB was named the AWS Technology Partner of the Year at the AWS Partner Summit Taiwan. Support for VPC peering Next, Atlas Stream Processing now supports VPC peering for self-hosted Apache Kafka on AWS and Amazon Managed Streaming for Apache Kafka (AWS MSK) . VPC peering is a secure method for connecting between virtual private clouds . As stream processing solutions like Atlas Stream Processing inherently connect to external data sources outside of MongoDB, the ability to make these connections as if your resources are on the same private network is a critical security requirement for many organizations. Users can select from any VPC peer configured within an Atlas project when setting up Kafka connections. Because peering is at the stream processing connection level, developers can configure Atlas Stream Processing to consume events from one Kafka cluster and produce them to another in a different VPC. Note that this feature has an additional cost. You can learn more in our documentation . Expanded support for Apache Kafka Third, we’re expanding capabilities for Apache Kafka in this release. Kafka is one of two key data sources Atlas Stream Processing supports today. One of Kafka’s strengths is its flexibility, allowing developers to customize configurations to suit various use cases, including those that rely on continuous stream processing. That flexibility can also create complexity, but Atlas Stream Processing focuses on making Kafka’s critical features easily accessible using the MongoDB Query API. By adding support for Kafka keys, developers can now read and write Kafka keys on their events, which enables filtering, partitioning, and aggregating based on key values. This ability provides greater control over routing processed data and is powerful for many stream processing use cases. Expanded Atlas Admin API support Lastly, we have added support for creating and deleting stream processors, as well as fetching operational stats of stream processors using the Atlas Admin API. Developers relying on the Admin API as a primary interface for interacting with Atlas will find this a welcome addition for managing their stream processors. Learn more in the documentation . With these new capabilities—additional AWS region support, VPC peering, the ability to use Kafka keys, and improved stream processing support for the Atlas Admin API—we've made it easier than ever for developers to integrate stream processing into their applications. We're excited to see the innovative ways you'll use these features. Ready to unlock the full potential of Atlas Stream Processing? Log in to Atlas today and start exploring the new features. We're eager to hear your feedback, so don't hesitate to share it with us on UserVoice . Your insights help us continue to improve and innovate.
Agnostiq & MongoDB: High-Performance Computing for All
Material scientists, computational biologists, and AI researchers all have at least one thing in common; the need for huge amounts of processing power, or ‘compute’, to turn raw data into results. But here’s the problem. Many researchers lack the skills needed to build the workflows that move data through huge networks of distributed servers, CPUs, and GPUs that actually do the number crunching. And that’s where Agnostiq comes in. Since the company’s inception in 2018, Agnostiq has put the power of high-performance computing (HPC) in the hands of researchers, bypassing the need for development expertise to build these essential data and compute pipelines. Power to the people “We started research on high-performance computing needs in fields like finance and chemistry, and through the process of onboarding, researchers quickly realized how hard it was for [researchers] to access and scale up on the cloud, or tap into HPC and GPU resources,'' said Santosh Kumar Radha, Agnostiq’s Head of Product. “If you wanted to scale up, there were not many tools available in the major cloud providers to do this.” To address this bottleneck, the team at Agnostiq built Covalent, a Python-based framework that allows researchers to easily design and run massive compute jobs on cloud platforms, on-prem clusters, and HPC services. With Covalent, startups and enterprises can build any AI or HPC application in a simple, scalable, and cost-effective way using a Python notebook, negating the need to interact with underlying infrastructure. One of the hardest challenges the Covalent team faced was combining traditional HPC with modern cloud technology. Because traditional HPC infrastructure was never designed to run in the cloud, the team spent considerable resources marrying techniques like GPU and CPU parallelization, task parallelization, and graph optimization with distributed cloud computing environments. As a result, researchers can use Covalent to quickly create a workflow that combines the convenience of cloud computing with specialized GPU providers and other HPC services. Everything, everywhere, all at once As the name suggests, Agnostiq has always focused on making their platform as open and resource neutral as possible. MongoDB Atlas , with its native multi-cloud capability, was a perfect complement. “At Agnostiq, everything we build has to be technology and vendor neutral. Interoperability is key for us,” said Radha. “We do all the mapping for our customers, so our platform has to perform a seamless transition from cloud to cloud.” The ability to move data between clouds became even more critical following the release of ChatGPT. With an explosion in generative AI research and development, the availability of GPU resources plummeted. “Resource scarcity in the ‘GPT era’ means you couldn’t get access to GPUs anywhere,” Radha added. “If you didn’t have a default cloud posture, you were nowhere, which is why we doubled down on multi-cloud and MongoDB Atlas to give our clients that optionality.” Open source opening doors Since the beginning, the team at Agnostiq has chosen MongoDB as their default NoSQL database. At first, the team adopted MongoDB’s free, open source product. “We didn’t have any DBAs as a small agile team. MongoDB gave us the freedom to build and manage our data workflows without the need for a specialist,” said William Cunningham, Head of HPC at Agnostiq. As their customer base grew along with the demand for cloud computing access, Agnostiq moved to MongoDB Atlas, gaining the freedom to move data seamlessly between AWS, Google Cloud, and Microsoft Azure. This gave Covalent the flexibility to reach multi-cloud compatibility at a faster rate than with standard tooling. Covalent provides a workflow management service by registering jobs, dispatching IDs, and collecting other metadata that allows fellow researchers and developers to reproduce the original work. MongoDB is used in the front-end, allowing a high volume of metadata and other assets to be published and cached in accordance with an event-driven architecture. This near real-time experience is key to a product aimed at delivering a unified view over distributed resources. MongoDB Atlas further provided the autoscaling required to grow with the user base and the number of workloads while keeping costs in check. “MongoDB Atlas helps us provide an ideal foundation for modern HPC and AI applications which require serverless compute, autoscaling resources, distributed workloads, and rapidly reconfigurable infrastructure,” added Radha. The future Looking to the future, Agnostiq is focused on servicing the huge demand for gen AI modeling and workflow building. To that end, the company released its own inference service called Function Serve within Covalent. Function Serve offers customers a complete, enterprise-grade solution for AI development and deployment, supporting serverless AI model training and fine-tuning. With Function Serve, customers can fine-tune, host, and serve any open-source or proprietary model with full infrastructure abstraction, all with only a few additional lines of code. MongoDB Atlas was used to rapidly develop a minimal service catalog while remaining cloud-agnostic. Looking ahead, the team plans to leverage MongoDB Atlas for enterprise and hybrid-cloud deployments in order to quickly meet customers in their existing cloud platforms. Agnostiq is a member of the MongoDB AI Innovators program , providing their team with access to Atlas credits and technical best practices. You can get started with your AI-powered apps by registering for MongoDB Atlas and exploring the tutorials available in our AI resources center . Additionally, if your company is interested in being featured, we'd love to hear from you. Reach out to us at ai_adopters@mongodb.com .
Meeting the UK’s Telecommunications Security Act with MongoDB
Emerging technologies like AI, IoT, and 5G have transformed the value that telecommunications companies provide the world. However, these new technologies also present new security challenges. As telcos continue to amass large amounts of sensitive data, they become an increasingly attractive target for cybercriminals — making both companies and countries vulnerable to cyberattacks. Fortunately, developers can protect user data which comes with strong security requirements on a developer data platform. By offering features to meet stringent requirements with robust operational and security controls, telcos can protect their customers’ private information. The UK Telecommunications Security Act Amid growing concerns about the vulnerability of telecom infrastructure, and its increasing digital dependency, the UK Telecommunications (Security) Act (TSA) was enacted on November 17, 2021. It was designed to bolster the security and resilience of the UK’s telecommunications networks. The TSA mandates that telecom operators implement rigorous security measures such as end-to-end encryption as well as identity and access management to protect their networks from a broad spectrum of threats, ensuring the integrity and continuity of critical communication services. The act allows the government to compel telecom providers to meet specific security directives. The United Kingdom’s Office of Communications (Ofcom) is a regulatory body responsible for overseeing compliance, conducting inspections, and enforcing penalties on operators that fail to meet the standards. The comprehensive code of practice included in the act offers detailed guidance on the security measures that should be implemented, covering risk management, network architecture, incident response, and supply chain security. The TSA tiering system The TSA establishes a framework for ensuring the security of public electronic communications networks and services. It categorizes telecoms providers into different tiers, with specific security obligations for each tier. The Act outlines three main tiers: Tier 1: These are the largest and most critical providers. They have the most extensive obligations due to their significant role in the UK's telecoms infrastructure. Tier 1 providers must comply with the full set of security measures outlined in the Act. Tier 2: These providers have a considerable role in the telecoms network but are not as critical as Tier 1 providers. They have a reduced set of obligations compared to Tier 1 but still need to meet substantial security requirements. Tier 3: These are smaller providers with a limited impact on the overall telecoms infrastructure. Their obligations are lighter compared to Tiers 1 and 2, reflecting their smaller size and impact. The specific obligations for each tier include measures related to network security, incident reporting, and supply chain security. The aim is to ensure a proportional approach to securing the telecoms infrastructure, with the highest standards applied to the most critical providers. Non-compliance may result in fines Under the TSA, non-compliance with security obligations can result in substantial fines. The fines are designed to be significant enough to ensure compliance and deter breaches. The significance of the fines imposed under the TSA underscores the importance the UK government places on telecom security and the serious consequences of failing to meet the established standards. How MongoDB can help MongoDB offers built-in security controls for all your data—whether your databases are managed on-premises with MongoDB Enterprise Advanced or with MongoDB Atlas , our fully managed cloud service. MongoDB enables enterprise-grade security features and simplifies deploying and managing your databases. Encrypting sensitive data The TSA emphasizes securing telecom networks against cyber threats. While specific encryption requirements are not detailed, the focus is on robust security practices, including encryption to protect data integrity and confidentiality. Operators must implement measures that prevent unauthorized access and ensure data security throughout transmission and storage. Compliance may involve regular risk assessments and adopting state-of-the-art technologies to safeguard the network infrastructure. MongoDB data encryption offers robust features to protect your data while it’s in the network, being stored, in memory, in transit (network), at rest (storage), and in use (memory, logs). Customers can use automatic encryption of key data fields like personally identifiable information (PII) or any data deemed sensitive—ensuring data is encrypted through its use. Additionally, with our industry-first Queryable Encryption , MongoDB offers a fast, searchable encryption scheme that supports equality searches, with additional query types such as range, prefix, suffix, and substring planned for future releases. Authentication and Authorization The TSA contemplates stringent identity and access management requirements to enhance network security. Regular audits and reviews of access permissions should be designed to prevent unauthorized access and to quickly identify and respond to potential security breaches. These measures aim to protect the integrity and confidentiality of telecommunications infrastructure. MongoDB enables users to authenticate to their Atlas UI with their Atlas credentials or via single sign-on with their GitHub or Google accounts. Atlas also supports MFA with various options, including OTP authenticators, push notifications, FIDO2 (hardware security keys or biometrics), SMS, and e-mail. MongoDB Enterprise Advanced users can authenticate to the MongoDB database using mechanisms including SCRAM, x.509 certificates, LDAP, OIDC, and passwordless authentication with AWS-IAM. Auditing Under the TSA, providers must implement logging mechanisms to detect and respond to security incidents effectively. Logs should cover access to sensitive systems and data, including unsuccessful access attempts, and must be comprehensive, capturing sufficient detail to facilitate forensic investigations. Additionally, logs should be kept for a specified minimum period and to be protected against unauthorized access, tampering, and loss. MongoDB offers granular auditing that monitors actions in your MongoDB environment and is designed to prevent and detect any unauthorized access to data, including CRUD operations, encryption key management, authentication, role-based access controls, replication, and sharding cluster operations. Additionally, MongoDB’s Atlas Organization Activity Feed displays select events that occurred for a given Atlas organization, such as billing or access events. Likewise, the Atlas Project Activity Feed displays select events that occurred for a given Atlas project. Network security The TSA outlines several network security requirements to ensure the protection and resilience of telecommunications networks. These requirements encompass various aspects of network security, including risk management, protection measures, incident response, and compliance with standards and best practices. Atlas offers many options to securely access your data with dedicated clusters deployed in a unique virtual private cloud (VPC) to isolate your data and prevent inbound network access from the internet. You can also allow a one-way connection from your AWS, Azure, or Google Cloud VPC/VNet to Atlas Clusters via Private Endpoints . Additionally, you can enable peering between your MongoDB Atlas VPC or VNet to your own dedicated application tier VPN with the cloud provider of your choice or enable only specific network segments to connect to your Atlas clusters via the IP Access list . In summary, the UK TSA is a critical regulatory framework aimed at protecting the nation’s telecommunications infrastructure from cyber threats. For telecom companies, compliance isn’t just a legal obligation but a business imperative. Failure to comply can mean significant financial penalties, reputational harm, and long-term operational challenges, underscoring the importance of adopting robust security measures and maintaining continuous adherence to the Act’s requirements. Visit MongoDB’s Strong Security Defaults page for more information on protecting your data with strong security defaults on the MongoDB developer data platform, as well as how to meet stringent requirements with robust operational and security controls.
Leveraging Database Observability at MongoDB: Real-Life Use Case
This post is the second in our three-part series, Leveraging Database Observability at MongoDB. Welcome back to the Leveraging Database Observability at MongoDB series. In our last discussion, we explored MongoDB's unique observability strategy using out-of-the-box tools designed to automatically monitor and optimize customer databases. These tools provide continuous feedback to answer critical questions such as what is happening, where is the issue, why is it occurring, and how do I fix it? This ensures enhanced performance, increased productivity, and minimized downtime. So let’s dive into a real-life use case, illustrating how different tools in MongoDB Atlas come together to address database performance issues. Whether you're a DBA, developer, or just a MongoDB enthusiast, our goal is to empower you to harness the full potential of your data using the MongoDB observability suite. Why is it essential to diagnose a performance issue? Identifying database bottlenecks and pinpointing the exact problem can be daunting and time-consuming for developers and DBAs. When your application is slow, several questions may arise: Have I hit my bandwidth limit? Is my cluster under-provisioned and resource-constrained? Does my data model need to be optimized, or cause inefficient data access? Do my queries need to be more efficient, or are they missing necessary indexes? MongoDB Atlas provides tools to zoom in, uncover insights, and detect anomalies that might otherwise go unnoticed in vast data expanses. Let’s put it into practice Let's consider a hypothetical scenario to illustrate how to track down and address a performance bottleneck. Setting the context Imagine you run an online e-commerce store selling a popular item. On average, you sell about 500 units monthly. Your application comprises several services, including user management, product search, inventory management, shopping cart, order management, and payment processing. Recently, your store went viral online, driving significant traffic to your platform. This surge increased request latencies, and customers began reporting slow website performance. Identifying the bottleneck With multiple microservices, finding the service responsible for increased latencies can be challenging. Initial checks might show that inventory loads quickly, search results are prompt, and shopping cart updates are instantaneous. However, the issue might be more nuanced and time-sensitive, potentially leading to a full outage if left unaddressed. The five-step diagnostic process To resolve the issue, we’ll use a five-step diagnostic process: Gather data and insights by collecting relevant metrics and data. Generate hypotheses to formulate possible explanations for the problem. Prioritize hypotheses to use data to identify the most likely cause. Validate hypotheses by confirming or disproving the top hypothesis. Implement and observe to make changes and observe the results. Applying the five-step diagnostic process for resolution Let’s see how this diagnostic process unfolds: Step 1: Gather Data and Insights Customers report that the website is slow, so we start by checking for possible culprits. Inefficient queries, resource constraints, or network issues are the primary suspects. Step 2: Generate Hypotheses Given the context, the application could be making inefficient queries, the database could be resource-constrained, or network congestion could be causing delays. Step 3: Prioritize Hypotheses We begin by examining the Metric Charts in Atlas. Since our initial check revealed no obvious issues, we will investigate further. Step 4: Validate Hypotheses Using Atlas' Namespace Insights , we break down the host-level measurements to get collection-level data. We notice that the transactions.transactions collection has much higher latency than others. By increasing our lookback period to a week, the latency increased just over 24 hours ago when customers began reporting slow performance. Since this collection stores details about transactions, we use the Atlas Query Profiler to find that the queries are inefficient because they’re scanning through the whole transaction documents. This validates our hypothesis that application slowness was due to query inefficiency. Figure 1: New Query Insights Tab Step 5: Implement and Observe We need to create an index to resolve the collection scan issue. The Atlas Performance Advisor suggests an index on the customerID field. Adding this index enables the database to locate and retrieve transaction records for the specified customer more efficiently, reducing execution time. After creating the index, we return to our Namespace Insights page to observe the effect. We see that the latency on our transactions collection has decreased and stabilized. We can now follow up with our customers to update them on our fix and assure them that the problem has been resolved. Conclusion By gathering the correct data, working iteratively, and using the MongoDB observability suite , you can quickly resolve database bottlenecks and restore your application's performance. In our next post in the "Leveraging Database Observability at MongoDB" series, we’ll show how to integrate MongoDB metrics seamlessly into central observability stacks and workflows. This 'plug-and-play' experience aligns with popular monitoring systems like Datadog, New Relic, and Prometheus, offering a unified view of application performance and deep database insights in a comprehensive dashboard. Sign up for MongoDB Atlas , our cloud database service, to see database observability in action. For more information, see our Monitor Your Database Deployment docs page.
Enhancing Retail with Retrieval-Augmented Generation (RAG)
In the rapidly evolving retail landscape, tech innovations are reshaping how businesses operate and interact with customers. Generative AI could add up to $275 billion of profit to the apparel, fashion, and luxury sectors’ by 2028, according to McKinsey analysis . One of the most promising developments in this realm is retrieval-augmented generation (RAG) , a powerful application of artificial intelligence (AI) that combines the strength of data retrieval with generative capabilities to supercharge retail enterprises. RAG offers compelling advantages specifically tailored for retailers looking to enhance their operations and customer engagement from personalization to enhanced efficiency. Let’s delve into how RAG is revolutionizing the retail sector. Check out our AI resource page to learn more about building AI-powered apps with MongoDB. Why RAG in retail Imagine a customer walks into your store, and based on their previous opt-in online interactions, your technology recognizes their preferences and seamlessly guides them through a personalized service—a feat made possible by RAG. Central to RAG’s effectiveness is its ability to integrate and analyze diverse data sources scattered across data warehouses. This integration enables retailers to gain comprehensive insights into their business performance, understand consumer behavior patterns, and make data-driven decisions swiftly. Below are some of the compelling advantages that RAG can offer: Personalization: RAG enables retailers to deliver highly personalized customer experiences by leveraging AI to understand and predict individual preferences based on past interactions. Operational efficiency: By integrating diverse data sources and optimizing processes like supply chain management, RAG helps retailers streamline operations, reduce costs, and improve overall efficiency. For example, RAG aids in tracking shipments and optimizing logistics—a traditional pain point in the industry. Data utilization: It allows retailers to harness the power of big data by integrating and analyzing disparate data sources, providing actionable insights for informed decision-making. Customer engagement: RAG facilitates proactive customer engagement strategies through features like autonomous recommendation engines and hyper-personalized marketing campaigns, thereby increasing customer satisfaction and loyalty. In essence, RAG empowers retailers to harness AI's full potential to deliver superior customer experiences, optimize operations, and maintain a competitive edge in the dynamic retail landscape. But without a clear roadmap, even the most sophisticated AI solutions can falter. By pinpointing specific challenges—such as optimizing inventory management or enhancing customer service—retailers can leverage RAG to tailor solutions that deliver measurable business outcomes. Despite its transformative potential, retailers must first be AI-ready and able to integrate it in a way that enhances operational efficiency without overwhelming existing systems. To achieve this, retailers need to address data silos, ensure data privacy, and establish robust ethical guidelines for AI use. According to a Workday Global Survey , only 4% of respondents said their data is fully accessible, and 59% say their enterprise data is somewhat or completely siloed. Without a solid data foundation, retailers will struggle to achieve the benefits they are looking for from AI. Embracing the future of retail with RAG and MongoDB By harnessing the power of data integration, precise use case definition, and cutting-edge AI technologies like RAG, retail enterprises can not only streamline operations but also elevate customer experiences to unprecedented levels of personalization and efficiency. Building a gen AI operational data layer (ODL) enables retailers to make the most of their AI-enabled applications. A data layer is an architectural pattern that centrally integrates and organizes siloed enterprise data, making it available to consuming applications. As shown below in Figure 1, pulling data into a single database eliminates data silos, centralizes data management, and improves data integrity. Using MongoDB Atlas to unify structured and unstructured operational data offers a cohesive solution by centralizing all data management in a scalable, cloud-based platform. This unification simplifies data management, enhances data consistency, and improves the efficiency of AI and machine learning workflows by providing a single source of truth. With a flexible data schema, retailers can accommodate any data structure, format, or source—which is critical for the 80% of real-world data that is unstructured . Figure 1: Generative AI data layer As AI continues to evolve, the retail industry is poised to see rapid advancements, driven by the innovative use of technologies like RAG. The future of retail lies in seamlessly integrating data and AI to create smarter, more responsive business models. If you would like to learn more about RAG for Retail, visit the following resources: Presentation: Retrieval-Augmented Generation (RAG) to Supercharge Retail Enterprises White Paper: Enhancing Retail Operations with AI and Vector Search: The Business Case for Adoption The MongoDB Solutions Library is curated with tailored solutions to help developers kick-start their projects Add vector search to your arsenal for more accurate and cost-efficient RAG applications by enrolling in the MongoDB and DeepLearning.AI course " Prompt Compression and Query Optimization " for free today.
Introducing AI-Powered Natural Language Mode in Atlas Charts
At MongoDB, our mission is to empower developers to seamlessly and efficiently build modern applications. To that end, we’ve announced a number of tools and improvements to help developers build faster , from AI-powered SQL query conversion in Relational Migrator to the MongoDB Provider for Entity Core Framework. In the same spirit, today at MongoDB.local Sydney we’re excited to announce the general availability of Natural Language Mode in Atlas Charts . Not only does this release help developers move faster with AI, it also showcases the work of MongoDB’s nimble Sydney engineering team. Intelligent developer experiences: The next level of productivity Over the past year, we’ve introduced a range of intelligent developer experiences across our platform, all of which aim to simplify and accelerate development processes. Overall, our goal is to make tools faster, better, and more connected for our developers, and to enable developers to leverage the power of AI to enhance their experience. As IBM highlights , true developer productivity involves delivering high-quality outputs that satisfy customers but also avoid developer burnout. By reducing learning curves, saving time, and providing easily accessible insights, intelligent features enable developers to focus on their most important work: building modern applications that solve real-world problems through outputs that are truly worth developers’ time and effort. Here’s what we’ve introduced: MongoDB Compass —Developers can use natural language to compose everything from simple queries to sophisticated, multi-stage aggregations. MongoDB Relational Migrator —With natural language, developers can convert SQL queries to MongoDB Query API syntax, streamlining migration projects. MongoDB Documentation —An intelligent chatbot, built on top of MongoDB Atlas and Atlas Vector Search, enables lightning-fast information discovery and troubleshooting during software development. By integrating AI into our most important developer tools, we’re helping developers cut through the noise and focus on creating innovative solutions. Simplifying data visualization with Natural language Mode Visualizing data can be an incredibly effective way of gleaning insights from application data, but creating effective visualizations can require specialized knowledge and experience with business intelligence (BI) tools. Atlas Charts was built to level the playing field. Now, with Natural Language Mode, developers can create visualizations simply by asking questions in plain English. Natural Language Mode reduces technical barriers and makes data visualization accessible to anyone with data in MongoDB Atlas. This means faster chart creation at an advanced scale—all within the Atlas ecosystem. Since announcing its development in the fall of 2023, we’ve made significant enhancements to Natural Language Mode, including an expanded suite of chart types built for all kinds of data from patient records to financial trading flows. We’ve also improved performance to ensure faster and more accurate chart generation, including the ability to handle more sophisticated prompts, filtering, sorting, binning, and limiting. In the next few weeks, we will add more chart variation, as well as another upgrade to our model that will increase accuracy and responsiveness by up to 50%. Natural Language Mode in action Now let’s walk through a few examples of what using Natural Language Mode in an Atlas Charts looks like: Generating a chart using Natural Language Mode With a simple query like "Show me the sales performance by country and product for Q4 FY2023," developers can instantly generate a relevant chart. Customizing charts in Classic Mode After generating a chart, developers can pull it into Classic Mode to fine-tune and customize it to fit their dashboard needs. Scheduling Dashboards Developers can also schedule their dashboards to be shared via email, ensuring that key stakeholders receive up-to-date insights automatically. More data, more insights Customers are already excited about the possibilities of Atlas Charts and Natural Language Mode, from operational analytics to embedded analytics. For example, one of MongoDB’s early customers has been using Natural Language Mode to track server performance across various regions. Business analysts leverage the feature to gain insights into server performance and share these insights internally. They plan to embed these visualizations into their customer portal using the embedding SDK offered by Atlas Charts . Another customer said: "As a developer with no prior experience in analytics, I was excited to see Natural Language Mode generate a clear value proposition that showcased what the product is capable of. It makes me want to throw more data in the database to get more insights." So check out Natural Language Mode in Atlas Charts today, and experience firsthand how AI can simplify and accelerate your data visualization workflows. Try out Natural Language Mode in Atlas Charts to transform your data visualization process. New to Charts? Register for MongoDB Atlas , deploy a cluster, and activate Charts for free.
Inside MongoDB’s Early Access Programs
In tech, staying ahead of the curve is essential. Tech companies must continually innovate and release new developments to meet the ever-evolving needs of users and global market demands. One successful strategy is the use of Early Access Programs (EAPs) and Preview Features. These initiatives offer unique opportunities for both companies and users, fostering mutually beneficial relationships that drive product excellence. You may have heard them referred to as "Beta Programs," "Pilot Programs," or "Feature Previews," but they all fall under the same category of early user engagement aimed at refining products before general release. So what are Early Access Programs and Preview Features? EAPs and Preview Features allow a select group of users to test and provide feedback on new features before general release. These programs are aimed at loyal customers, power users, or those who’ve shown interest in the company's products. EAPs grant users early access to upcoming features or products still in development. Participants engage closely with the product team, providing valuable insights and feedback that shape the final product. Preview Features are specific features in a product made available for users to try before the official release. Unlike beta testing, which may involve a full product, preview features are often isolated components of a larger system. What are the Benefits of Early Access Programs? They offer numerous advantages for both companies and users. Direct feedback from real users helps identify bugs, usability issues, and areas for improvement that may not be evident during internal testing. This actionable feedback is crucial in enhancing product quality, allowing companies to release more polished and reliable features. This means higher user satisfaction and fewer post-release problems. EAPs also increase customer engagement and appreciation among participants by offering a hands-on exclusive experience to them. This can foster loyalty and strengthen the relationship between the company and its customers. These programs also provide market validation, offering an opportunity to gauge market interest and demand for new features. This enables companies to make data-driven decisions about which features to prioritize and further invest in. By identifying and resolving potential issues early in the development process, EAPs reduce the risk of major failures post-launch and allow for better resource allocation and planning. How does MongoDB approach Early Access Programs? At MongoDB, EAPs are classified into two categories: private preview and public preview. Private preview entails an invite-only white-glove experience for a select few participants who closely test and provide feedback. Public preview implies the feature is available to the public to try, either as downloadable tools or features in Atlas. Key Elements Selective Participation: MongoDB’s EAPs are typically invitation-only, targeting power users, those who’ve shown significant interest in new features, or are part of our MongoDB Champions community. This selective approach ensures that feedback comes from experienced users who can provide valuable insights. Direct Collaboration: Participants engage directly with MongoDB’s product and engineering teams. This direct line of communication allows for real-time feedback, in-depth discussions, and a collaborative approach to feature development. Structured Feedback Collection: We use a variety of methods to collect feedback, including surveys, structured interviews, and feedback forms. This structured approach ensures that feedback is actionable and can be effectively integrated into the development process. Iterative Development: Feedback is used to make iterative improvements to the features being tested. This agile approach allows us to quickly address issues, refine functionalities, and enhance the overall user experience. Transparency and Updates: We maintain open and transparent communication with our EAP participants, providing regular updates on the status of the features, changes made based on their feedback, and future development plans. This transparency fosters trust and keeps participants engaged throughout the program. Rewards and Recognition: Participants can share their stories on our social channels, be part of global events, and win MongoDB swag! Be Part of MongoDB’s Innovation MongoDB’s Early Access Programs offer participants the chance to gain early access to innovative features, influence product development, and join an exclusive, engaged community. As we continue to innovate and expand our product offerings, our Early Access Programs will help us deliver high-quality, user-centric solutions that empower our customers to achieve their strategic objectives. Join us in shaping the future of MongoDB by enrolling in our early access programs today!
Welcome to the (Tech) Olympics!
Welcome to the Tech Olympics, where code meets competition! With the 2024 Summer Olympics starting today, we thought it’d be fun to imagine developers as athletes, showcasing their skills in a series of thrilling events. From relay races to coding challenges, the Tech Olympics would bring together the brightest minds in tech for a competition like no other. Whether you're a coding wizard, a bug-squashing maestro, or an AI aficionado, there would be something to test your limits and celebrate your talents. Opening ceremony The opening ceremony is one of the most iconic aspects of the Olympics. From the lighting of the torch to performances by local artists, the opening ceremony encapsulates the spectacle of the games, and is a necessity for the Tech Olympics. The Tech Olympics opening ceremony would kick off with a grand procession of teams involved, adorned in attire representing their area of expertise. Next, there’d be a performance by artists and developers using augmented and virtual reality to blend art with cutting-edge technology. Finally, there would be the lighting of the torch, but instead of the flame being run across the country, an application would be written and passed between developers from around the world that, when run, would light the torch and start the games. Now that we’ve kicked off the Tech Olympics, let's consider what its events might look like. Code sprint relay The "code sprint relay" would be a collaborative coding event where teams of developers would tackle a series of coding challenges in relay format. The twist would be that each member could only code for a set period (say, 5-10 minutes) before handing the code off to the next person. This setup requires clear communication and strategic planning, as each coder must quickly understand and build upon their predecessor's work. Code sprint relay challenges would range from algorithm problems to debugging tasks, demanding various skills and swift adaptability. This event would be fast-paced and dynamic, with a lively atmosphere filled with the buzz of coding and quick exchanges of ideas. Success would be measured not only by the completion of challenges but also by the efficiency and quality of the code, making this event a test of teamwork and technical skill under pressure. Security capture the flag Capture the flag might seem more like a kids’ game than an Olympic event, but trust us, there’d be nothing childish about this event. The "security capture the flag" event would be an exciting cybersecurity competition in which participants would need to solve security-related challenges to capture hidden "flags." These challenges would range from web application exploits and reverse engineering, to cryptographic puzzles and network forensics. Working in teams, participants would race against the clock to uncover vulnerabilities, exploit them, and find the embedded flags within a controlled, simulated environment. At the end, a debriefing session would highlight the most innovative solutions and techniques used. Success would be measured by the number of flags captured and the ingenuity of the approaches, showcasing participants' technical skills and strategic thinking under pressure. Bug hunt Have you ever built out your code and then, upon running it, realized that you made a mistake? If you have, you’ll understand just how intense this next event could be! The "bug hunt challenge" would be a fast-paced competition in which participants are tasked with finding and fixing bugs within a complex codebase. Each individual would be given the same software project with numerous hidden bugs, ranging from simple syntax errors to intricate logical flaws. Participants must use their debugging skills and tools to identify and resolve as many issues as possible within a set time limit. The event would be marked by intense focus and strategic problem-solving as competitors meticulously comb through the code. An automated system would verify the fixes instantly, ensuring accuracy and efficiency. Success would be measured by the quantity and severity of bugs resolved, along with the quality of the fixes, making this event a test of attention to detail and technical proficiency. AI arena We’d be remiss not to include an AI event! The "AI arena" event would be a competitive showcase where participants create machine learning models using a provided dataset to solve a specified problem. Teams would have several hours to analyze the data, create features, and train their models. The objective would be to develop a model with the highest accuracy and performance, balancing technical innovation with practical application. In the end, teams would present their models to judges, explaining their methodologies and challenges faced. Judging criteria would include model accuracy, creativity, and clarity of the presentation, making this event a comprehensive test of technical and communication skills. Location Finally, you can’t have an Olympics without a city to host it. There are plenty of tech hubs to choose from—San Francisco, London, Beijing—but we thought it’d be more fun to pick a growing tech hub like Ha Noi, Vietnam, as our location. Vietnam had the highest digital economy growth in Southeast Asia in 2022 , putting it on the path to be named alongside other “tech giant” cities. Also, Vietnamese food is excellent! During the games, local startups and tech companies would showcase their work on the world stage, and visiting developers would see the innovations that Vietnamese companies are working on. Sadly, there won’t be an actual Tech Olympics this year, but maybe in the future, there will be. An event that will bring the world's best developers together to showcase their skills, foster friendly competition, and allow the world to see just how amazing developers are. If you have some ideas about other events you would want to see at a Tech Olympics, connect with us on X (Twitter) and let us know what your ideas are. Interested in learning more about or connecting more with MongoDB? Join our MongoDB Community to meet other community members, hear about inspiring topics, and receive the latest MongoDB news and events.
Building Gen AI Applications Using Iguazio and MongoDB
AI can lead to major enterprise advancements and productivity gains. By offering new capabilities, they open up opportunities for enhancing customer engagement, content creation, process automation, and more. According to McKinsey & Company, generative Al has the potential to deliver an additional $200-340B in value for the banking industry . One popular use case is customer service, where gen AI chatbots have quickly transformed the way customers interact with organizations. They handle customer inquiries and provide personalized recommendations while empathizing with them and offering nuanced support tailored to individual needs. Another less obvious use case is fraud detection and prevention. AI offers a transformative approach by interpreting regulations, supporting data cleansing, and enhancing the efficacy of surveillance systems. These systems can analyze transactions in real-time and flag suspicious activities more accurately, which helps institutions prevent monetary losses. In this post, we introduce the joint MongoDB and Iguazio gen AI solution which allows for the development and deployment of resilient and scalable gen AI applications. Before diving into how it works and its value for you, let’s first discuss the challenges enterprises face when operationalizing gen AI applications. Check out our AI resource page to learn more about building AI-powered apps with MongoDB. Challenges to operationalizing gen AI Building an AI application starts with a proof of concept. However, enterprises need to successfully operationalize and deploy models in production to derive business value and ensure the solution is resilient. Doing so comes with its own set of challenges such as: Engineering challenges - Deploying gen AI applications requires substantial engineering efforts from enterprises. They need to maintain technological consistency throughout the operational pipeline, set up sufficient infrastructure resources, and ensure the availability of a team equipped with a comprehensive ML and data skillset. Currently, AI development and deployment processes are slow, time-consuming, and fraught with friction. LLM risks - When deploying LLMs, enterprises need to reduce privacy risks and comply with ethical AI standards. This includes preventing hallucinations, ensuring unbiased outputs, filtering out offensive content, protecting intellectual property, and aligning with regulatory standards. Glue logic and standalone solutions - The AI landscape is vibrant, and new solutions are frequently being developed. Autonomously integrating these solutions can create overhead for ops and data professionals, resulting in duplicate efforts, brittle architectures, time-consuming processes, and a lack of consistency. Iguazio and MongoDB together: High-performing and simplified gen AI operationalization The joint Iguazio and MongoDB solution leverages the innovation of these two leading platforms. The integrated solution allows customers to streamline data processing and storage, ensuring gen AI apps reach production while eliminating risks, improving performance, and enhancing governance. MongoDB for end-to-end AI data management MongoDB Atlas , an integrated suite of data services centered around a multi-cloud NoSQL database, enables developers to unify operational (structured and unstructured data), analytical, and AI data services into a single platform to streamline building AI-enriched applications . MongoDB’s flexible data model enables easy integration with different AI/ML platforms, allowing organizations to adapt to changes in the AI landscape without extensive infrastructure modifications. MongoDB meets the requirements of a modern AI and vector data store: Operational and unified: MongoDB’s ability to serve as the operational data store (ODS) enables financial institutions to efficiently handle large volumes of real-time operational data and unifies AI/vector data, ensuring AI/ML models use the most accurate information. It also enables organizations to meet compliance and regulatory requirements (e.g., 3DS2, ISO20022, PsD2) by the timely processing of large data volumes. Multi-modal: Alongside structured data, there's a growing need for semi-structured and unstructured data in gen AI applications. MongoDB's JSON-based multi-modal document model allows you to handle and process diverse data types, including documents, network/knowledge graphs, geospatial data, and time series data. Atlas Vector Search lets you search unstructured data. You can create vector embeddings with ML models and store and index them in Atlas for retrieval augmented generation (RAG), semantic search, recommendation engines, dynamic personalization, and other use cases. Flexible: MongoDB’s flexible schema design enables development teams to make application adjustments to meet changing data requirements and redeploy application changes in an agile manner. Vector store: Alongside the operational data store, MongoDB serves as a vector store with vector indexing and search capabilities for performing semantic analysis. To help improve gen AI experiences with greater accuracy and mitigate hallucination risks, using a RAG architecture together with the multi-modal operational data typically required by AI applications. Deployment flexibility: MongoDB can be deployed self-managed on-premise, in the cloud, or in a SaaS environment. Or deployed across a hybrid cloud environment for institutions not ready to be entirely on the public cloud. Iguazio’s AI platform Iguazio (acquired by McKinsey) is an AI platform designed to streamline the development of ML and gen AI applications in production at scale. Iguazio’s gen AI-ready architecture includes capabilities for data management, model development, application deployment, and LiveOps. The platform—now part of QuantumBlack Horizon , McKinsey’s suite of AI development tools—addresses enterprises’ two biggest challenges when advancing from gen AI proofs of concept to live implementations within business environments. Scalability: Ensures uninterrupted service regardless of workload demands, scaling gen AI applications when required. Governance: Gen AI guardrails mitigate risk by directing essential monitoring, data privacy, and compliance activities. By automating and orchestrating AI, Iguazio accelerates time-to-market, lowers operating costs, enables enterprise-grade governance, and enhances business profitability. Iguazio’s platform includes LLM customization capabilities, GPU provisioning to improve utilization and reduce cost, and hybrid deployment options (including multi-cloud or on premises). This positions Iguazio to uniquely answer enterprise needs, even in highly regulated environments, either in a self-serve or managed services model (through QuantumBlack, McKinsey’s AI arm). Iguazio’s AI platform provides: Structured and unstructured data pipelines for processing, versioning, and loading documents. Automated flow of data prep, tuning, validating, and LLM optimization to specific data efficiently using elastic resources (CPUs, GPUs, etc.). Rapid deployment of scalable real-time serving and application pipelines that use LLMs (locally hosted or external) as well as the required data integration and business logic. Built-in monitoring for the LLM data, training, model, and resources, with automated model re-tuning and RLHF. Ready-made gen AI application recipes and components. An open solution with support for various frameworks and LLMs and flexible deployment options (any cloud, on-prem). Built-in guardrails to eliminate risks and improve accuracy and control. Examples: Building with Iguazio and MongoDB #1 Building a smart customer care agent The joint solution can be used to create smart customer care agents. The diagram below illustrates a production-ready gen AI agent application with its four main elements: Data pipeline for processing the raw data (eliminating risks, improving quality, encoding, etc.). Application pipelines for processing incoming requests (enriched with data from MongoDB’s multi-modal store), running the agent logic, and applying various guardrails and monitoring tasks. Development and CI/CD pipelines for fine-tuning and validating models, testing the application to detect accuracy risk challenges, and automatically deploying the application. A monitoring system collecting application and data telemetry to identify resource usage, application performance, risks, etc. The monitoring data can be used to improve the application performance further through an RLHF (reinforcement learning from human feedback) integration. #2 Building a hyper-personalized banking agent In this example, accompanied by a demo video , we show a banking agent based on a modular RAG architecture that helps customers choose the right credit card for them. The agent has access to a MongoDB Atlas data platform with a list of credit cards and a large array of customer details. When a customer chats with the agent, it chooses the best credit card for them, based on the data and additional personal customer information, and can converse with them in an appropriate tone. The bank can further hyperpersonalize the chat to make it more appealing to the client and improve the odds of the conversion, or add guardrails to minimize AI hallucinations and improve interaction accuracy. Example customer #1: Olivia Olivia is a young client requesting a credit card. The agent looks at her credit card history and annual income and recommends a card with low fees. The tone of the conversation is casual. When Olivia asks for more information, the agent accesses the card data while retaining the same youthful and fun tone. Example customer #2: Miss Jessope The second example involves an older woman who the agent calls “Ms Jessope”. When asking for a new card, the agent accesses her credit card history to choose the best card based on her history. The conversation takes place in a respectful tone. When requesting more information, the response is more informative and detailed, and the language remains respectful. How does this work under the hood? As you can see from the figure below, the tool has access to customer profile data in MongoDB Atlas collection bfsi.user_data and is able to hyperpersonalize its response and recommendations based on various aspects of the customer profile. A RAG process is implemented using the Iguazio AI Platform with MongoDB Atlas data platform. The Atlas Vector Search capabilities were used to find the relevant operational data stored in MongoDB (card name, annual fees, client occupation, interest rates, and more) to augment the contextual data during the interaction itself to personalize the interaction. The virtual agent is also able to talk to another agent tool that has a view of the credit card data in bfsi.card_info (such as card name, annual and joining fees, card perks such as cashback, and more), to pick a credit card that would best suit the needs of the customer. To ensure the client gets the best choice of card, a guardrail is added that filters the cards chosen according to the data gathered by the agent as a built-in component of the agent tool. In addition, another set of guardrails is added to validate that the card offered suits the customer by comparing the card with the optimal ones recommended for the customer’s age range. This whole process is straightforward to set up and configure using the Iguazio AI Platform, with seamless integration to MongoDB. The user only needs to create the agent workflow and connect it to MongoDB Atlas, and everything works out of the box. Lastly, as you can see from the demo above, the agent was able to leverage the vector search capabilities of MongoDB Atlas to retrieve, summarize, and personalize the messaging on the card information and benefits in the same tone as the user’s. For more detailed information and resources on how MongoDB and Iguazio can transform your gen AI applications, we encourage you to apply for an exclusive innovation workshop with MongoDB's industry experts to explore bespoke modern app development and tailored solutions for your organization. Additionally, you can enjoy these resources: Start implementing gen AI applications in your enterprise today How Leading Industries are Transforming with AI and MongoDB Atlas The MongoDB Solutions Library is curated with tailored solutions to help developers kick-start their projects Add vector search to your arsenal for more accurate and cost-efficient RAG applications by enrolling in the MongoDB and DeepLearning.AI course " Prompt Compression and Query Optimization " for free today.