Skip to content

Sellwyse Project Dashboard

AI-powered health platform with wearable integration, Web3 rewards, and Telegram growth engine.


What We're Building

Sellwyse is a health & fitness ecosystem that connects BLE wearable devices (rings, bands, gym equipment) to mobile apps, a coaching platform, and a web dashboard — all powered by AI analytics and blockchain rewards.

Products

Product Type Stack Description
User App (app-user) Mobile Flutter (iOS + Android) End-user app — syncs wearables, tracks health metrics (heart rate, SpO2, sleep, steps), AI coaching, food photo analysis, workout logging
Device App (app-device) Mobile Flutter (iOS + Android) Dedicated BLE device manager — firmware updates, raw data viewer, device pairing/config
Coach App (coach-app) Mobile Flutter (iOS + Android) For coaches/trainers — view client metrics, assign workout plans, real-time video sessions (Agora), AI-assisted recommendations
Champion Stats Hub (champion-stats-hub) Web React / Vite / TypeScript Web dashboard — leaderboards, analytics charts, Solana NFT badge gallery, admin views
Telegram Mini App (telegram-mini-app) Web (embedded) React / Vite / TypeScript Viral growth engine — health challenges inside Telegram, TON token rewards, share cards, referral system
Backend API (backend-api) Server Rust / Axum / PostgreSQL Core API — auth, device sync, health data pipeline, AI orchestration, Solana minting, Telegram bot
Athion SDK (athion-sdk) Library Dart Shared SDK — BLE protocols, data models, API client, used by all Flutter apps

Tracks Overview

Track 1: MVP (242 tasks)

Core platform build-out. Device sync & deduplication, health metrics pipeline, user/coach/device apps, web dashboard, real-time video coaching, push notifications, AI coaching via Gemini Pro, TTS voice feedback via ElevenLabs. Touches all repos.

Track 2: Early Release (81 tasks)

Early release app enhancements — UI revamp across all apps + deep AI analytics. Personalized health scores, trend detection, anomaly alerts, food photo analysis (Gemini Vision), AI-generated workout plans, DeepSeek fallback for bulk processing. Touches backend-api, athion-sdk, app-user, app-device, coach-app, champion-stats-hub.

Track 3: Solana / Web3 (48 tasks)

Blockchain integration. Compressed NFT achievement badges via Bubblegum, health-to-earn token design, Solana wallet linking, NFT gallery on web dashboard. Touches backend-api, athion-sdk, app-user, champion-stats-hub.

Track 4: Telegram Mini App (45 tasks)

Viral acquisition channel. Health challenges (steps, hydration, sleep), TON token rewards, leaderboards, share cards, referral tracking, deep-links to main app. Touches telegram-mini-app, backend-api.

Track 5: Marketing & Growth (44 tasks)

Launch strategy. Landing page, App Store listings, build-in-public content, influencer outreach, community building, analytics setup. Primarily content/marketing work, touches landing page and store listings.


Track Progress

Track Tasks Status Deadline
Track 1: MVP 242 Not started Mar 20, 2026
Track 2: Early Release 81 Not started May 1, 2026
Track 3: Solana / Web3 48 Not started Apr-Jun 2026
Track 4: Telegram Mini App 45 Not started May-Jul 2026
Track 5: Marketing 44 Not started Ongoing
Research 14 Not started Ongoing
Total 474

Total Monthly Budget Summary

All estimates include 2x safety buffer.

During Development (Now — May 2026)

Item Monthly
MiniMax AI (M2.5 High-Speed) $80
Track 1 MVP (Mac + ngrok + APIs) $110-264
Track 5 Marketing (Canva, domain) $26-70
Dev Phase Total $216-414/mo

Production — Early (10 users, Track 1 only)

Item Monthly
MiniMax AI $80
Track 1 MVP (AWS) $96-244
Track 5 Marketing $26-476
Early Prod Total $202-800/mo

Production — Full (1K users, all tracks live)

Track 2 replaces Track 1 infrastructure at scale. Tracks 3-5 are additive.

Item Monthly
MiniMax AI $80
Track 1+2 (AWS + AI analytics) $370-1,000
Track 3 Solana (additional) +$14-174
Track 4 Telegram 1K users (additional) +$54-164
Track 5 Marketing +$26-476
Full Prod Total (1K users) $544-1,894/mo

Production — Scale (100K Telegram users)

Item Monthly
MiniMax AI $80
Track 1+2 (AWS + AI analytics) $370-1,000
Track 3 Solana +$14-174
Track 4 Telegram 100K users +$288-1,280
Track 5 Marketing +$26-476
Scale Prod Total (100K users) $778-3,010/mo

One-Time Costs

Item Cost
Track 3 Merkle tree (Solana) $100
Track 5 (logo, App Store accounts, domain) $268-378
One-Time Total $368-478

Development relies on AI coding agents to execute the 474 tasks across all tracks. Help with MiniMax AI payment is needed to keep development velocity up.


Monthly Cloud Costs (2x Safety Buffer)

Track Dev Phase Production
Track 1: MVP $110-264/mo $96-244/mo (10 users)
Track 2: Early Release included $370-1,000/mo (1K users)
Track 3: Solana +$14-174/mo
Track 4: Telegram +$54-1,280/mo (1K-100K)
Track 5: Marketing $26-70/mo $26-476/mo

Devices

Device Type Status
X3B Ring BLE wearable ring Integrated
JVC Band BLE wearable band Integrated
Godzilla Station Serial gym equipment Integrated
Gorillas BLE wearable (TBD) Arriving ~Mar 30

Repos

Repo Stack Tracks
backend-api Rust / Axum / PostgreSQL 1, 2, 3, 4
athion-sdk Dart 1, 2, 3
app-user Flutter 1, 2, 3
app-device Flutter 1, 2
coach-app Flutter 1, 2
champion-stats-hub React / Vite / TypeScript 1, 2, 3
telegram-mini-app React / Vite / TypeScript 4 (new)

How Task Tracking Works

Each task has its own .md file with YAML frontmatter:

---
id: "BE-001"
status: todo        # todo | in_progress | done
started_at:         # filled when agent starts
completed_at:       # filled when agent completes
---

When an agent completes a task, it updates the task file's status to done and sets completed_at. Push to trigger Amplify rebuild and the dashboard updates automatically.

File locations: docs/tasks/track-{n}/{TASK-ID}.md