Corporate RAG platform on terabytes of data
AI Platform · RAG · Semantic Search · Document Intelligence
The client
Bell Group, a large UK construction and facilities services company with more than 2,500 employees and extensive corporate archives distributed across multiple storage systems.
Challenge
Employees had to search across terabytes of fragmented documents and archives, and finding the right file took up to 15 minutes. The company needed a centralized knowledge layer that could provide fast semantic search while preserving existing user access permissions.
Scope & features
- Corporate knowledge platform
- RAG-based document retrieval
- Semantic search
- Centralized archive indexing
- Natural-language document search
- Source-linked answers
- Role-based access control
- Enterprise document integration
- Knowledge onboarding support
Solution
Built a centralized RAG platform that indexed corporate archives and lets employees search company knowledge by meaning rather than keywords, with direct links to original sources.
Outcome
Document search time dropped from ~15 minutes to seconds — roughly 150× faster — with better access to internal knowledge and faster employee onboarding.
Tech stack
- Python
- Node.js
- OpenAI API
- LangChain
- PostgreSQL
- Pinecone
- AWS S3
- React
≈150×
faster search: 15 min → seconds
Start with an AI audit
A focused review of your operations, data and automation — and a prioritized roadmap with ROI estimates. The same entry point our long-term partners started from.
Talk to the team