The Challenge
The raw material sat as audio files and Word documents in SharePoint with no ingestion pipeline, no transformation logic and no quality controls. The service had to work in Welsh and English from the first day, and it had to be safe enough for people in active withdrawal. In a consumer product a poor answer costs someone a bad afternoon. In a stop smoking service commissioned by a health board, a response that invents coping advice or mishandles a disclosure of risk is a safeguarding incident.
Our Strategy
We transcribed 30,236 recorded coaching calls into a structured corpus of expert interventions, each carrying the trigger context, the coaching strategy and the technique. Claude does the synthesis under a narrow instruction: carry the clinical substance of the retrieved record exactly, express it in the coaching voice, invent nothing. The safeguarding layer is what allowed any of it to be deployed. Risk detection runs before anything is generated, ahead of the knowledge base search, so on detection the system stops before it writes a word, and never produces an answer it then has to check. A human is alerted, the event is recorded inside the client's own environment, and Welsh speakers receive the same safety response in Welsh. Refusing to continue is a harder guarantee than handling it well.
The Results
30236
recorded coaching calls distilled into 44K+ expert interventions
364
people supported since Joules went live in February 2026
90%
first-time pass on the response evaluation suite, 125 of 139 questions
Testimonial
" Unloq has been a great partner to work with. Their approach helped us identify where AI should be applied and the AI solution they developed is now delivering on the needs of our client and their users. "
- Stuart Jackson, CEO & Founder, ICE Creates
