Building AI Applications with MongoDB

Choosing the Right Open-Source Blueprint

• Define the use case - understand the goals and data needs
• Identify relevant open-source blueprints
• Ensure the blueprints are well-documented with comprehensive guides and tutorials
• Verify the quality of the documentation and community support
• Ensure the blueprint is well-integrated with MongoDB and the ecosystem
• Consider security and regulatory compliance for the industry

The MongoDB AI Applications (MAP) Program

• Launched to help customers build and deploy modern AI applications
• Combines MongoDB technology with leading AI partners and expert services
• Aims to accelerate time-to-market, minimize risks, and maximize value of AI investments
• Provides access to the most advanced AI technologies through strong partnerships
• Offers professional services like strategic advisory, demos, and integrations
• Addresses challenges like data integration, evolving AI landscape, and lack of in-house expertise

Retrieval Augmented Generation (RAG) Example

• Demonstrates a healthcare use case for breast cancer prevention and diagnosis
• Leverages MongoDB's document model to handle structured, semi-structured, and unstructured data
• Uses MongoDB change streams to keep data updated and integrated
• Generates reports and conclusions using generative AI, reducing a 6-month process to 10 minutes
• Allows natural language queries to search and retrieve relevant information
• Integrates with MongoDB's Vector search and Atlas search capabilities

Key Takeaways

• MongoDB's AI Applications (MAP) program provides a comprehensive solution to build and deploy modern AI applications
• MongoDB's document model, Vector search, and other capabilities make it well-suited for AI workloads
• The Retrieval Augmented Generation (RAG) example showcases how MongoDB can power real-world, meaningful AI applications
• Attendees can explore the free training offered to learn how to build generative AI applications on MongoDB and AWS