Upload and Query
- 1 Upload documents via API gateway
- 2 Documents chunked and embedded automatically
- 3 Ask natural language questions
- 4 Get sourced answers with citations
DocIntel
In ProgressEnterprise Document Intelligence
A 10-component distributed document intelligence system built with Kotlin API gateway, Python Haystack RAG pipeline, and full Docker orchestration. Demonstrates production-grade patterns: hybrid search with RRF fusion, multi-tenant isolation, semantic caching, and LiteLLM provider abstraction.
// screenshots
// features
What it does
Hybrid Search with RRF Fusion
Combines BM25 keyword search and vector similarity for superior retrieval quality.
Multi-Tenant Document Isolation
Document-level access control via Qdrant payload filtering with pre-filter ACLs.
Semantic Caching
Qdrant dual-use for vector storage and query caching, reducing LLM costs.
LiteLLM Provider Abstraction
Same code for Ollama (local) or any cloud LLM provider — no vendor lock-in.
Langfuse Observability
Self-hosted tracing and monitoring for the entire RAG pipeline.
// architecture
Under the hood
10-component distributed system with Kotlin API gateway, Python RAG pipeline, and Docker orchestration.
API Gateway
Kotlin/Spring Boot
REST API, authentication, rate limiting
Document Service
Kotlin/Spring Boot
Document ingestion, chunking, processing
RAG Service
Python/Haystack + LiteLLM
Query pipeline with hybrid retrieval and reranking
Admin Service
Kotlin/Spring Boot
Tenant management and system administration
Qdrant
Vector DB
Vector storage and semantic cache
PostgreSQL + pgvector
Database
Metadata storage and fallback vector search
// user journeys
How it gets used
// tech stack
HaystackKotlinQdrantLiteLLMDocker
🧠
Phoenix
Ready when you are
Hey! I'm Phoenix — I know Titas's work, projects, and experience. Ask me anything — from distributed systems to production RAG, or what it's like building at Tesco and VMware.