AskBase: RAG-Powered Internal Knowledge Assistant
Generative AI / RAG
Project Overview
An internal Q&A assistant built on retrieval-augmented generation (RAG) — letting employees query a company's internal knowledge base, SOPs, and documentation in natural language, with source-cited answers.
Client: Concept Build — Harfield Operations Group
Duration: 5 weeks
The Challenge
Employees were spending significant time searching through SharePoint folders, PDF SOPs, and lengthy Word documents to find answers to routine operational questions. New starters in particular needed weeks to get up to speed.
Our Solution
Documents (PDFs, Word files, Confluence exports) are ingested via an async Python pipeline that chunks, embeds using a sentence-transformer model, and stores vectors in Pinecone. At query time, LangChain retrieves the top-k relevant chunks and passes them as context to Claude via the Anthropic API, which generates a cited, conversational answer referencing the source document. The React frontend shows the answer alongside expandable source citations so users can verify. Deployed via Docker with GitHub Actions managing continuous delivery.
Results
- Average time to find SOPs and policy answers reduced from 15+ minutes to under 60 seconds
- New starter onboarding score improved significantly in post-deployment survey
- Source citations in every answer maintained trust and allowed quick verification
- Ingestion pipeline processes new or updated documents automatically on upload
Client Testimonial
"Alicorn built the LiteCloud practice management platform for us — it handles our entire client workflow, task tracking and reporting. It has been running in production for over 8 months and the team has been responsive throughout. The co-founders are directly involved, which makes a real difference."
Lee Phillips
Digital Data Lead, Twinings Ovaltine
Ready to Start Your Project?
Let's discuss how we can bring your ideas to life and drive your business forward.
Get in Touch