Partnered on product direction where “helpful AI” had to stay trustworthy for stressed users: clear states, recovery paths, and human-visible boundaries around what the system can and cannot do.
Created product and technical direction for bringing generative and agentic AI into a native mobile experience, including multi-agent decomposition, scoped prompts, and supervised routing between specialist capabilities.
Worked across product, UX, and implementation so the team could test faster in market without losing sight of production realities.
- Created a roadmap for integrating generative and agentic AI into the mobile product, with staged delivery rather than a big-bang rebuild.
- Broke down broad caregiver problems into narrower specialist functions, routed by a lead/supervisor layer to keep outputs more focused and explainable.
- Helped redesign the experience from v1 to v2 so the product felt clearer, calmer, and more trustworthy for real users.
- Supported integration between the mobile app and web/API-based AI workflows so ideas could move from prototype to usable product faster.
- Outcome: faster feedback loops, clearer product direction, and a stronger bridge between founder vision, UX, and what could actually be released.
Build with Etherly 




