Hacknarok

A dashboard for daily psychophysical state analysis — you record a short video with your camera and microphone, and the app shows how you feel today and how it's been changing over time.

How it works

1. You open the dashboard and record a ~2-minute clip (camera + microphone).
2. The backend runs the recording through an analysis pipeline: video (facial expression, fatigue), audio (voice) and text (what you're saying).
3. The result is a single daily score plus a breakdown into a few sub-scores.
4. Results are saved to your history and shown as a calendar heatmap, a trend chart and a detailed per-day summary.

Stack

• Backend: Python + FastAPI
• Frontend: Next.js 16 + React 19 + TailwindCSS v4

Repo layout

• backend/ — API, analysis pipeline and result storage.
• frontend/ — browser dashboard (recording + visualizations).

Running locally

You'll need Python and Node.js. Start the backend first.

Backend (port 8000)

cd backend
python -m venv .venv
# Windows:
.venv\Scripts\activate
# macOS/Linux:
source .venv/bin/activate

pip install -r requirements.txt
python -m app.seed           # generate initial history
uvicorn app.main:app --reload

Swagger UI: http://localhost:8000/docs.

Frontend (port 3000)

cd frontend
npm install
npm run dev

Dashboard: http://localhost:3000.

If the backend runs on a different host, set it in frontend/.env.local:

NEXT_PUBLIC_API_URL=http://localhost:8000

Status

Built for a hackathon — the end-to-end skeleton (recording, API, scoring, history, UI) is fully working, but some ML models in the pipeline are still mocked and waiting to be swapped for real implementations. Details in the code.