YDB background

AI capabilities in YDB

YDB for AI: vector search

Capabilities

  • Exact vector search
  • Approximate vector search without an index
  • Approximate vector search with an index
  • More details in the documentation

Use cases

  • Recommendation systems
  • Semantic search
  • Similar image search
  • Anomaly detection
  • Classification systems

Solves business problems

  • Information search based on freely formulated natural language queries
  • Search in unstructured data: images, video, and audio
  • Improves quality of recommendations and personalization

AI for YDB: agents, MCP servers, frameworks

MCP server for YDB

  • Analyze table structures with neural networks
  • Generate and debug programs working with YDB
  • More details in the documentation

NeuroAdmin

  • Debug performance issues in chat mode
  • Available in Enterprise Manager in Yandex Cloud

LangChain framework

  • Use LangChain to develop applications based on LLMs
  • Seamlessly connect YDB as a vector store using langchain-ydb integration
  • More details in the documentation

Vector search in YDB helps build AI applications

Accelerates similarity search and reduces database load

Improves search engine accuracy through semantic similarity

Facilitates big data analysis and pattern detection