HIPAA-compliant conversational AI for a UK healthcare company
A multi-agent conversational platform for patients, built for a regulated environment where every answer has to be traceable.
The challenge
Spryt needed conversational AI that patients could rely on, in a setting where data handling, auditability and accuracy are regulated. Off-the-shelf chatbots could not show where an answer came from or prove what happened to patient data.
What I did
- Architected HIPAA-compliant infrastructure with end-to-end data traceability and logging.
- Designed multi-agent conversational pipelines with Rasa, LangGraph and retrieval-augmented generation (RAG), with compliance guardrails built into the flow rather than added afterwards.
- Built structured logging and monitoring so every conversation could be audited.
- Delivered high-availability cloud infrastructure for production traffic.
Technology
More case studies
Architecture and technical roadmap for four concurrent AI platforms
Set the technical direction for four AI products at once: a real-time voice toolkit, public-sector agent systems, and a compliance-aware financial agent for Australia.
15+ conversational AI systems across healthcare, finance, retail and education
Ran a small AI services practice end to end, from first call to production, for clients in eight countries.
Product strategy and pivot for a no-code conversational AI startup
Took a no-code conversational AI platform from idea to MVP, validated it with customers, and pivoted to where the demand was.
Planning something similar?
Tell me what you're trying to achieve. I reply within 24 hours with a suggested next step.