80% of customer support digitized
A RAG-powered AI chatbot integrated into the Peps Digital platform, answering PSDM users' questions directly from the interface, 24/7.
- Client
- Peps Digital
- Sector
- SaaS · Healthcare (PSDM)
- Year
- 2025
Key results
80%
Questions auto-answered
92%
Satisfaction rate
1,200+
Questions per month
Context
Peps Digital builds a SaaS platform for PSDMs (French home healthcare providers) and medical equipment vendors. Their support team was completely underwater, overwhelmed by recurring usage questions from the platform's customers. Advisors spent most of their day answering level 1 requests at the expense of strategic work.
Challenge
Let platform users get an immediate answer to their usage questions, 24/7, directly from the interface. With precision high enough to handle sometimes complex workflows, without replacing humans on the topics that deserve them.
Our approach
AI chatbot natively integrated into the Peps Digital platform interface. RAG architecture (Retrieval-Augmented Generation): the product's full documentation is indexed in OpenAI's Vector Store and queried on every question. The chatbot returns a contextual answer with relevant doc excerpts, and escalates to a human when the question falls outside its scope.
Deliberately minimalist stack: OpenAI for the LLM and Vector Store, Next.js for the entire platform. No separate backend, no infra to maintain. Fewer moving parts, faster deployments and iterations.
Result
Chatbot deployed in the platform, used directly by PSDMs. 80% of support requests are now handled without human intervention, according to Bruno. The team freed up to focus on product guidance and strategic clients.
Tech stack
- OpenAI
- Vector Store
- RAG
- Next.js
Let's build together
Give your teams backtheir Fridays.
We listen. We analyze. We build. With you.
