The project
Mobile app aiming for CE DMN certification. Tracks mental load continuously via multimodal analysis (wearables, calendar, notifications, self-assessments, journal) processed by an on-device Small Language Model. On-device first: no raw data leaves the phone, the Go backend is "health-data-free" by design.
Highlights
- On-device first: local SLM computes the score
- Go backend that never sees health data in clear
- Multi-agent architecture (collection, analysis, interaction, export)
- Free premium for under-18 users (equity + youth focus)
Stack
FlutterGollama.cppONNX RuntimeSQLCipherHealthKit / Health Connect
Product, architecture, regulatory framing.
My role on this project.
Contact
Let's talk?
Easiest route: a 30-minute call to understand your context. Email works just as well.