Remove Fly.io and GitHub Actions workflows

- Delete all GitHub Actions workflow files
- Remove Fly.io deployment configuration references
- Update TASKS.md and implementation plan to reflect Hugging Face deployment
- Clean up stale CI/CD documentation

.github/workflows/deploy-backend.yml

@@ -1,61 +0,0 @@
-name: Deploy Backend → Fly.io
-
-on:
-  push:
-    branches: [main]
-    paths:
-      - 'backend/**'
-      - 'Dockerfile'
-      - '.dockerignore'
-      - 'fly.toml'
-      - 'Mediapipe_XGBoost/**'
-      - 'CNN_Autoencoder_LightGBM/**'
-      - 'CNN_PreTrained/**'
-
-jobs:
-  test:
-    name: Run backend tests
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-
-      - uses: actions/setup-python@v5
-        with:
-          python-version: '3.12'
-          cache: pip
-          cache-dependency-path: backend/requirements.txt
-
-      - name: Install deps
-        run: |
-          pip install -r backend/requirements.txt
-          pip install -r backend/requirements-dev.txt
-
-      - name: Run tests
-        working-directory: backend
-        env:
-          KERAS_BACKEND: tensorflow
-          TF_CPP_MIN_LOG_LEVEL: "3"
-          CUDA_VISIBLE_DEVICES: ""
-          TF_ENABLE_ONEDNN_OPTS: "0"
-        run: pytest tests/ -v --tb=short -q
-        # Note: tests will be skipped automatically if model .pkl files are absent
-        # (model artefacts are gitignored). Add them as GitHub Actions artifacts
-        # or use DVC/GCS to restore them in CI if you want full test coverage.
-
-  deploy:
-    name: Deploy to Fly.io
-    needs: test
-    runs-on: ubuntu-latest
-    environment: production
-    concurrency:
-      group: fly-deploy
-      cancel-in-progress: true
-    steps:
-      - uses: actions/checkout@v4
-
-      - uses: superfly/flyctl-actions/setup-flyctl@master
-
-      - name: Deploy
-        run: flyctl deploy --remote-only
-        env:
-          FLY_API_TOKEN: ${{ secrets.FLY_API_TOKEN }}

.github/workflows/deploy-frontend.yml

@@ -1,45 +0,0 @@
-name: Deploy Frontend → Vercel
-
-on:
-  push:
-    branches: [main]
-    paths:
-      - 'frontend/**'
-
-jobs:
-  build-and-deploy:
-    name: Build & Deploy
-    runs-on: ubuntu-latest
-    environment: production
-    steps:
-      - uses: actions/checkout@v4
-
-      - uses: actions/setup-node@v4
-        with:
-          node-version: '22'
-          cache: npm
-          cache-dependency-path: frontend/package-lock.json
-
-      - name: Install dependencies
-        working-directory: frontend
-        run: npm ci
-
-      - name: Type-check
-        working-directory: frontend
-        run: npx tsc --project tsconfig.app.json --noEmit
-
-      - name: Build
-        working-directory: frontend
-        env:
-          VITE_WS_URL:  ${{ vars.VITE_WS_URL }}
-          VITE_API_URL: ${{ vars.VITE_API_URL }}
-        run: npm run build
-
-      - name: Deploy to Vercel
-        uses: amondnet/vercel-action@v25
-        with:
-          vercel-token:   ${{ secrets.VERCEL_TOKEN }}
-          vercel-org-id:  ${{ secrets.VERCEL_ORG_ID }}
-          vercel-project-id: ${{ secrets.VERCEL_PROJECT_ID }}
-          working-directory: frontend
-          vercel-args: '--prod'

SanketSetu_ Production-Grade Implementation Plan.md

@@ -11,9 +11,9 @@ The system follows a modern microservices-inspired pattern to ensure scalability
 | Component | Technology | Role | Hosting (Free Tier) |
 | :--- | :--- | :--- | :--- |
 | **Frontend** | React + Vite + TS | User interface, webcam capture, real-time feedback | **Vercel** |
-| **Backend API** | FastAPI (Python) | WebSocket management, API gateway, logic | **Fly.io** |
-| **Inference Engine** | ONNX Runtime / XGBoost | High-speed model execution | **Fly.io** (Internal) |
-| **Storage** | Cloudflare R2 | S3-compatible storage for model weights | **Cloudflare** |
+| **Backend API** | FastAPI (Python) | WebSocket management, API gateway, logic | **Hugging Face Spaces** |
+| **Inference Engine** | ONNX Runtime / XGBoost | High-speed model execution | **Hugging Face Spaces** |
+| **Storage** | Hugging Face Model Hub | Model weights and assets | **Hugging Face** |
 | **Real-time** | WebSockets (WSS) | Low-latency frame-by-frame data transfer | N/A |
 
 ---
@@ -58,16 +58,12 @@ The frontend is designed to be "cool," responsive, and highly interactive, provi
 
 ## 4. Deployment & DevOps
 
-### 4.1 Continuous Integration/Deployment (CI/CD)
-Using **GitHub Actions**, the project will follow a strict deployment pipeline:
-1.  **Lint & Test**: Ensure code quality and run unit tests for ML logic.
-2.  **Build**: Create optimized production builds for the React app and Dockerize the FastAPI backend.
-3.  **Deploy**: 
-    *   Frontend automatically pushes to **Vercel**.
-    *   Backend pushes to **Fly.io** using `flyctl`.
+### 4.1 Deployment Strategy
+1.  **Frontend**: Manually push to **Vercel** using Vercel CLI or GitHub integration (when needed).
+2.  **Backend**: Manually deploy to **Hugging Face Spaces** using Git push or Hugging Face Hub CLI.
 
 ### 4.2 Scalability & Cost Management
-*   **Scale-to-Zero**: The backend on Fly.io can be configured to sleep when not in use to preserve free-tier resources.
+*   **Auto-scaling**: Hugging Face Spaces manages resource allocation automatically with free-tier CPU instances.
 *   **CDN Caching**: Vercel's Edge Network will cache all static assets, ensuring fast load times globally.
 
 ---
@@ -95,5 +91,5 @@ Using **GitHub Actions**, the project will follow a strict deployment pipeline:
 [1] [FastAPI Documentation](https://fastapi.tiangolo.com/) - High-performance web framework for building APIs.
 [2] [MediaPipe Hands](https://developers.google.com/mediapipe/solutions/vision/hand_landmarker) - Real-time hand landmark detection.
 [3] [Framer Motion](https://www.framer.com/motion/) - A production-ready motion library for React.
-[4] [Fly.io Free Tier](https://fly.io/docs/about/pricing/) - Details on free-tier resource allocation.
+[4] [Hugging Face Spaces](https://huggingface.co/docs/hub/spaces) - Free-tier hosting for ML applications.
 [5] [Vercel Deployment](https://vercel.com/docs/deployments/overview) - Global CDN and hosting for frontend applications.

TASKS.md

@@ -40,7 +40,7 @@ SanketSetu/
 │   ├── weights/                ← Symlink or copy of model pkl files
 │   ├── requirements.txt
 │   ├── Dockerfile
-│   └── fly.toml
+
 │
 ├── frontend/                   ← Vite + React + TS
 │   ├── src/
@@ -217,25 +217,16 @@ SanketSetu/
 - [x] Add `.dockerignore` (excludes `.venv`, `node_modules`, `*.pth`, tests)
 - [ ] Test locally: `docker build -t sanketsetu-backend . && docker run -p 8000:8000 sanketsetu-backend`
 
-### 3.2 Fly.io Configuration
-- [x] Create `fly.toml` (repo root, region=maa, port 8000, shared-cpu-2x)
+### 3.2 Hugging Face Spaces Configuration
+- [x] Create Hugging Face Spaces repository for backend deployment
 - [x] Note: Keras/TF will increase Docker image size — use `tensorflow-cpu` to keep slim
-- [ ] Set secrets via `flyctl secrets set` for any API keys
-- [ ] Run: `flyctl deploy --dockerfile Dockerfile`
+- [ ] Push Docker image to Hugging Face Container Registry
 
 ### 3.3 Vercel Frontend Deployment
 - [x] Create `frontend/vercel.json` with SPA rewrite + WASM Content-Type header
 - [x] Add `VITE_WS_URL` and `VITE_API_URL` to Vercel environment variables (via CI vars)
 - [ ] Ensure `@mediapipe/tasks-vision` WASM files are served correctly (add to `public/`)
 
-### 3.4 GitHub Actions CI/CD
-- [x] Create `.github/workflows/deploy-backend.yml`
-  - Triggers on push to `main` when `backend/**` changes
-  - Steps: checkout → setup Python → run tests → `flyctl deploy`
-- [x] Create `.github/workflows/deploy-frontend.yml`
-  - Triggers on push to `main` when `frontend/**` changes
-  - Steps: checkout → `npm ci` → tsc → `npm run build` → Vercel CLI deploy
-
 ---
 
 ## Phase 4 — Testing & Hardening