Top 10% of Women Selected, Lovable SheBuilds Hackathon
Built an AI buddy that turns distance to presence
275 million people worldwide face 167 unsupported hours per week during recovery, where 40-60% relapse. I built YANA for a loved one in recovery living far away, realizing the real problem isn't structured process like therapy, it's the isolation between sessions.
Key Insight: Existing apps track sobriety (restriction-based). Behavioral science shows you must replace habits, not just stop them. The gap: No tool helps people fill unstructured hours with purpose and structure like yana
Target: Adults in recovery needing 24/7 support during the 167 hours structured help doesn't cover.
Started with a PRP - Product Requirements Prompt. The PRP defines the initial query and context given to generative AI to produce the PRD. It is the brief before the brief.
The PRP forced me to think about what I’m really asking AI to do and whether the context is strong enough. If I don’t think deeply, how can AI know what to build?
Next, was the PRD - to define the goals, features, user flow and success. Lastly, I built using Lovable.dev.
Built end-to-end in under 48 hours during the SheBuilds Hackathon, YANA demonstrated the ability to go from idea to functional prototype quickly with clear real-world utility.
Users described interactions with YANA as “empathetic,” “non-judgmental,” and “encouraging,” validating the value of human-centered AI in sensitive scenarios.