Faber is an AI-powered fitness application featuring a chatbot coach, workout generator, body scanner, and recipe creator. The app is built and functional but has the following specific issues that need to be resolved:
1. Body Scanner — Accuracy & Consistency
Results are consistently 5–6% too high on body fat percentage. Additionally, the same scan under identical conditions produces 1–2% variation between scans. We need the accuracy improved and results made consistent. We are also open to your recommendation on whether switching from live scanning to photo upload would improve accuracy.
2. Workout Generator — Logic & Sync
Exercises outside the selected muscle group are being included in generated workouts. Sets, duration, and other parameters are also out of sync between the app and the admin panel. We need a full audit of the workout generation logic and admin panel synchronisation.
3. AI Chatbot — Database Integration
The chatbot is not retrieving and applying data from our knowledge database correctly. It gives surface-level answers even when the user asks for depth, and fails to distinguish between related topics properly. The answers exist in the database — the issue is in how the chatbot queries and uses that data. We need a proper investigation and proposed fix.
We need someone who can review the existing codebase, identify root causes, and deliver clean fixes. Please include in your proposal your relevant experience, your initial thoughts on these issues, and your estimated timeline and cost.
Apply tot his job
Apply To this Job