Skip to content

Latest commit

 

History

History
22 lines (19 loc) · 1.62 KB

File metadata and controls

22 lines (19 loc) · 1.62 KB

TODO

Improvements

  • Add structured tracing around Graphiti operations (ingestion/query) with optional OpenTelemetry exporters.
  • Add memory-efficient streaming for large document processing to handle files >100MB without memory issues.
  • Implement connection pooling and health checks for Neo4j database connections.
  • Add data validation and sanitization (guardrails) for user inputs to prevent injection attacks and malformed data.
  • Implement rate limiting for API calls to external services to respect usage quotas.
  • Add performance monitoring and metrics collection for ingestion and query operations.
  • Switch to other LLM models when default one is down.
  • Implement caching layer for frequently accessed graph data and embeddings.

New Features

  • Implement additional document processors for .csv, .pptx, .xlsx, and remote URLs with automatic metadata enrichment.
  • Introduce a background ingestion pipeline (queue/worker) to support large batch uploads with progress tracking.
  • Offer multi-provider LLM support with pluggable clients (Gemini, OpenAI, Azure, Anthropic) and smart fallback handling.
  • Add graph visualization capabilities with export to various formats (PNG, SVG, JSON).
  • Add search with advanced filtering options (date ranges, content types, relevance scores).
  • Implement knowledge graph analytics dashboard showing insights, popular topics, and projects available.
  • Implement knowledge graph export/import functionality for backup and migration purposes.
  • Add support for interactive query refinement with follow-up questions and context narrowing.