Contact person: Business Client

Phone:Show

Email:Show

Location: Kochi, India

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"Android AI Fitness Coach App — Real-Time Polar Heart-Rate + Audio Coaching + ChatGPT Backend




Summary

I want to build a modular Android fitness-training app that starts as my personal Heavy-Duty workout assistant and can later be released commercially.
The app should read real-time heart-rate data from Polar (Verity Sense / H10) via Bluetooth LE, give in-set audio coaching through any Bluetooth headset, log sets automatically, and generate a post-workout summary + next-session plan using ChatGPT (via a central backend API).
Later I’ll monetize through ads or subscriptions.

Objectives

1. Work flawlessly for my personal use first (offline/local).
2. Architect it so it can later:
3. show AdMob ads or premium unlocks,
4. connect to a central OpenAI relay server (my key, not user’s),
5. handle multiple users safely.

Core Features (v1 Personal Build)

A. Real-Time Coaching

Connect to any BLE Heart-Rate Service (0x180D) device.
Primary test devices: Polar Verity Sense / Polar H10.

Stream HR in real time (1 Hz).

Local logic (no latency):

Detect HR plateau → cue “Approaching failure – one controlled rep.”

End set → start 60 s timer; at 60 s compare HR drop; cue “Extend rest 30 s” or “Ready for next set.”

Text-to-speech audio to any Bluetooth headset (Android TTS).

Offline; works without network during workout.


B. Workout Logging

Voice or quick-type input (e.g. “Leg press 350×6 RPE 9”).

Auto-capture HR peak + duration + RPE.

Save locally (Room/SQLite).


C. Post-Workout Summary

On “End Workout,” send summary JSON to my backend endpoint (OpenAI Chat Completions call).

Receive concise text:

2-line performance summary

Load/rest recommendation

“Next session in X days.”


Show summary on History screen.


D. Morning Readiness

Pull sleep / resting HR / HRV proxy via Health Connect / Google Fit.

Compute readiness label: TRAIN / OPTIONAL / REST.



Architecture Requirements

Layer Tech Note

App Kotlin + Jetpack Compose Local UI + BLE + TTS + DB
Data Room (offline) CSV export
Backend Lightweight Node.js or Firebase Function Calls OpenAI Chat API using my key
Ads (future) AdMob hooks prepared Disabled for personal build
Config API base URL + model name stored in remote JSON for easy change


Commercial Upgrade Path

1. v1: private APK, no ads, my OpenAI key in backend.


2. v1.1: enable AdMob + subscription toggle.


3. v2: multi-user login (Firebase Auth), central analytics dashboard.



All design and code must be my property so I can extend or publish later.

Ownership & Rights

I own 100 % of source code, assets, IP, and backend.

Developer may not reuse or resell.

Deliverables include full Android Studio project, keystore, and backend source.


Deliverables

1. Android APK + source with live BLE, TTS cues, Health Connect readiness, local DB.


2. Backend API (Node or Firebase) with one endpoint:
/ai/summarizeWorkout → { summary, nextSessionDays }


3. Documentation: pairing instructions, permission list, build/run steps.


Milestone

Week Task

1 Compose UI + logging + Room
2 BLE HR stream + TTS cues
3 Health Connect + Readiness
4 OpenAI relay backend + integration + polish
Bonus Stable build + on-time


Acceptance Tests

1. Connects to Polar Verity Sense/H10; HR updates ≤ 2 s latency.


2. In-set cue triggers near failure based on HR threshold.


3. Rest-gate cue works at 60 s.


4. Voice input logs correctly.


5. OpenAI summary returns within 5 s after workout.


6. Readiness card updates daily from Health Connect.


7. No crashes; battery impact minimal (< 5 %/hr).




What to Include in Your Bid

The word Mentzer in the first line.

Link/APK of a BLE heart-rate app you’ve built (preferably Polar/0x180D).

One paragraph on how you’ll handle low-latency in-set cues (local vs server).

Confirmation you’ve used Health Connect or Google Fit before.

Preferred stack (Kotlin Compose or Flutter) and timeline (≤ 4 weeks).


Budget

Fixed price ₹50000 INR (max).
Payments via milestone escrow on Freelancer/Upwork.



Notes

Must work offline except for post-workout AI call.

Backend OpenAI key stays server-side only.

Future versions will use the same architecture for multi-user subscriptions.

I’ll provide: Polar device, OpenAI API key, test phone, and sample data." (client-provided description)