Ground it in the source
An answer should point back to the regulation, guideline or SOP it came from. That's why EmMA retrieves from EMA documents instead of relying on memory.
I'm Rob Jones: TMF specialist, member of the TMF Reference Model Steering Committee, and 15+ years across CROs, sponsors and vendors. I write about using AI where quality, compliance and inspection readiness matter, and share what I build with it along the way.

AI is moving faster than our guidance. These are the rules I hold myself to while the rest catches up.
An answer should point back to the regulation, guideline or SOP it came from. That's why EmMA retrieves from EMA documents instead of relying on memory.
AI drafts, sorts and suggests. A qualified person still reviews and decides, and the record shows who did what.
If it touches GCP data it gets the computerised-system treatment: intended use, testing, evidence. AI makes that faster, not optional.
Know where prompts and documents go. Pocket TMF collects no data at all, and that's the default I start from.
Everything here was built with AI as my pair programmer. Pocket TMF is a native app; EmMA is AI all the way down.

The TMF and ISF Reference Models on your phone, searchable and fully offline, on iOS and Android.
30+countries
700+downloads


AI · RAGA chatbot grounded in EMA regulations, guidelines and inspection findings. Ask it about the TMF.
Chat with EmMAThe hardware, software and AI tools I use to design and build everything here.
See the setup

Two free collections for phone and desktop, made with AI.
Browse wallpapersProject management, regulatory affairs and clinical operations, now focused on the Trial Master File. If you're working on clinical trials, AI or healthcare innovation, I'd love to hear from you.