Stellanet is an AI-assisted research outreach platform that helps students find relevant faculty and prepare higher-quality first-contact emails.
The product combines grounded discovery, transparent match reasoning, and controllable draft rewriting in a single workflow.
- Discover professor candidates from real publication and affiliation signals.
- Rank matches based on user research interest and profile context.
- Show trust details such as why-match bullets, score breakdowns, and recent paper snippets.
- Provide contact email signals when available (direct, inferred, or unavailable).
- Generate outreach drafts and rewrite by tone while keeping human review in control.
- Support authentication flows including signup/signin, email verification, and password reset.
- Support Google OAuth sign-in.
- Set your interest, profile, and target universities.
- Review ranked faculty matches with reasoning signals.
- Generate and refine draft outreach.
- Send only after manual review.
- Frontend: React, Vite, Tailwind CSS
- Backend: FastAPI, Python
- Model layer: Amazon Nova (Bedrock)
- Research data source: OpenAlex
- Auth and session store: SQLite (local) or PostgreSQL (production)
- Deployments: Vercel (frontend), AWS App Runner (backend)
frontend/Stellanet- frontend applicationbackend- API, retrieval, ranking, rewrite, auth, and mailer logic
cp backend/.env.example backend/.env
cp frontend/Stellanet/.env.example frontend/Stellanet/.envcd backend
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
uvicorn main:app --reload --host 127.0.0.1 --port 8000cd frontend/Stellanet
npm install
npm run dev -- --host 127.0.0.1 --port 5173- Set
VITE_API_BASE_URLin frontend env to your backend base URL. - Set Bedrock credentials and model env vars in backend env for Nova-powered generation.
- Set
OPENALEX_MAILTOfor OpenAlex requests. - For contact form delivery, set
CONTACT_TO_EMAILand either Resend or SMTP env vars. - For Google sign-in, set frontend
VITE_GOOGLE_CLIENT_IDand backendGOOGLE_OAUTH_CLIENT_IDS.
- Improve ranking calibration and confidence transparency.
- Add saved workspaces and outreach history.
- Add stronger observability and automated quality checks.
- Expand personalization controls for draft generation and rewriting.