Home/AI Innovation/RAG Infrastructure

RAG Infrastructure

http://localhost:3336
Open ↗

Click to open live demo

http://localhost:3336

Key Capabilities

  • ChromaDB vector store with all-MiniLM-L6-v2 sentence-transformer embeddings
  • 75+ indexed knowledge articles with semantic chunking and metadata enrichment
  • Sub-100ms retrieval with relevance scoring and contextual re-ranking
  • Dual-provider sentiment analysis layered on retrieval: VADER (5ms) + DistilBERT (50-200ms)
  • Reusable across use cases: agent assist, knowledge search, reconciliation, Q&A
  • FastAPI backend with client-side LRU caching (50 entries, 60s TTL)

Overview

Production-grade Retrieval Augmented Generation pipeline using ChromaDB vector store, sentence-transformer embeddings, and intelligent chunking. Indexes enterprise knowledge bases and serves contextual results in under 100ms. The foundation layer that powers NEXUS, CX AI agent assist, and knowledge search across demos.

By the Numbers

Retrieval<100ms
Embeddings384-dim
Knowledge Base1000s of Docs

Tech Stack

ChromaDBSentence TransformersFastAPIall-MiniLM-L6-v2VADER + DistilBERT