All work
AI
RAG assistantKnowledge Assistant
A grounded internal assistant that answers staff questions from trusted documentation.
Year
2025
Timeline
4 months
Team
4 engineers
Focus
RAG assistant
Results that matter
- −60%
- faster time-to-answer
- 92%
- staff adoption
- 100%
- answers with citations
The challenge
Staff at a healthcare provider spent hours hunting through scattered policies, procedures and guidelines, and couldn't always trust what they found.
Our approach
We built a retrieval-augmented assistant grounded strictly in approved sources, with a citation on every answer, role-based access control, and an evaluation harness that catches regressions before each release.
The outcome
Time-to-answer fell 60%, 92% of staff adopted it within a quarter, and every response is traceable to its source, so the team genuinely trusts it.
What we delivered
Answers grounded in approved sources
Citation on every response
Role-based access control
Continuous evaluation harness
Tech stack
PythonAnthropicLangChainPostgreSQLRedisNext.js
“It's the first internal tool people actually enjoy using. The citations are what won everyone over, they trust the answers.”
, Director of Operations, Healthcare provider
Next case study
Demand ForecastingDemand forecasting that cut stockouts and waste across a regional network.
Have a project in mind?
Tell us about it, we'll get back to you within 48 hours.