Product
Roadmap
RailQueryIQ
AI-powered conversational access to railroad technical documentation at the wayside.
Contact us about RailQueryIQ →Overview
RailQueryIQ brings retrieval-augmented generation to railroad technical documentation. Manuals, site plans, wiring configurations, and railroad-specific policies become conversationally queryable at the wayside, without dependence on back-office connectivity.
The architecture uses lightweight embedding and quantized small-LLM inference (Phi-3-mini or Llama 3.2 3B at the time of design) for edge deployment. RailQueryIQ is on the OEG roadmap.
Key features
- → Unified searchable document repository (vendor manuals, site plans, wiring configs, policies)
- → RAG query engine with natural-language understanding
- → Multi-document cross-device queries (e.g., SEAR-to-signal path tracing)
- → Site-specific configuration integration for context-aware answers
- → Edge deployment with optional back-office document feeds
Technologies & Protocols
RAG Quantized LLM (Phi-3-mini / Llama 3.2) Apache Tika pdf.js Tesseract OCR PostgreSQL + pgvector Sparkplug B
Roadmap
This product is in development and not yet available. Contact us to discuss your timeline or requirements.