Depth First Social (DFS)

Connect through shared thinking. Match with people who think like you using your Gemini conversation history — no swiping, no quizzes. Built at DeerHacks 2026.

---

What it does

• Upload Gemini data — Export your Gemini Apps Activity (JSON). We build a 6-axis interest profile (Cognitive, Emotional, Collaboration, Communication, Motivation, Lifestyle) and never store raw messages.
• Find matches — Set your intent (Hackathon partner, Friend, Co-founder, etc.) and get ranked matches via vector similarity (Snowflake).
• Chat — Open a private chat with a match. Chats expire after 10 minutes or when both users end the conversation; messages and match rows are deleted.
• Token gallery — Each time you update your Gemini data you earn a token. Your gallery shows every update; clicking a token opens your spider chart.

---

Tech stack

Layer	Stack
Frontend	Next.js 16 (App Router), React 19, Auth0, Supabase (profiles, matches, messages, Realtime), Tailwind
Backend	FastAPI, Python 3 — extract/scrub Gemini JSON, Snowflake connector, optional Solana minting
Data	Supabase: profiles, matches, messages. Snowflake: user archetype vectors, cosine similarity / Cortex Search
Auth	Auth0 → profile sync to Supabase (first_name, last_name, avatar, etc.)
Optional	Solana (devnet) for wallet connect; Gemini API for match blurbs

---

Repo structure

deerhacks/
├── frontend/          # Next.js app (dashboard, matchmaker, chat, token gallery)
├── backend/           # FastAPI (Gemini extraction, /v2/archetype, Snowflake upsert)
├── solana-program/    # Anchor program (reputation / soulbound identity)
├── demo/              # Demo Gemini JSON + setup script for test users
├── docs/              # Architecture (Gemini → Snowflake pipeline)
└── frontend/supabase/migrations/   # SQL for profiles, matches, messages, upload_history

---

Setup

1. Frontend (Next.js)

cd frontend
cp .env.example .env
# Edit .env: Auth0, Supabase, Snowflake, Gemini, NEXT_PUBLIC_API_URL
npm install
npm run dev

Runs at http://localhost:3000.

Required env (see .env.example):

• Auth0 — AUTH0_SECRET, AUTH0_DOMAIN, AUTH0_CLIENT_ID, AUTH0_CLIENT_SECRET, APP_BASE_URL
• Supabase — NEXT_PUBLIC_SUPABASE_URL, NEXT_PUBLIC_SUPABASE_ANON_KEY, SUPABASE_SERVICE_ROLE_KEY
• Backend URL — NEXT_PUBLIC_API_URL (e.g. http://localhost:8000)
• Snowflake (for matching) — SNOWFLAKE_ACCOUNT, SNOWFLAKE_USER, SNOWFLAKE_PASSWORD, SNOWFLAKE_WAREHOUSE, SNOWFLAKE_DATABASE, SNOWFLAKE_SCHEMA
• Gemini (match blurbs) — GEMINI_API_KEY

2. Backend (FastAPI)

cd backend
pip install -r requirements.txt
cp .env.example .env
# Edit .env: GEMINI_API_KEY, Snowflake vars
uvicorn main:app --reload --port 8000

Runs at http://localhost:8000. Used for archetype vector upsert (POST /v2/archetype) and optional Gemini extraction.

3. Supabase

1. Create a project at supabase.com.
2. Run the migrations in SQL Editor (or via CLI) from frontend/supabase/migrations/:
   • profiles (id, email, first_name, last_name, avatar_url, wallet_address, sol_balance, karma_score, archetype_scores, archetype, summary, upload_history, etc.)
   • matches (id, user_a, user_b, match_blurb, status, closed_by_user_a, closed_by_user_b, expires_at)
   • messages (id, match_id, sender_id, content, created_at)
   • Add upload_history (JSONB) and match expiry/close columns if not present (see migration filenames).

4. Snowflake (matching)

• Create a database/schema and tables (e.g. USER_ARCHETYPES with user_id, server_id, vector column).
• Backend and frontend use the same Snowflake credentials for embedding upsert and match search (cosine similarity or Cortex Search).
• See docs/ARCHITECTURE_GEMINI_JSON_AND_VECTORS.md and frontend/lib/snowflake.ts for the flow.

---

Running the app

1. Start backend: cd backend && uvicorn main:app --reload --port 8000
2. Start frontend: cd frontend && npm run dev
3. Open http://localhost:3000 → Sign up / Log in (Auth0) → Upload Gemini JSON → Generate chart → Find matches → Open chat.

---

Features in the UI

• Home — Landing, nav (logo, Wallet, Log out / Log in, Sign up), hero, testimonials, CTA.
• Dashboard — Profile card (avatar, name, tokens link), Upload Data / My Chart tabs, Your Tokens gallery (each token opens the chart), Find My Match (modal with search and Connect → chat).
• Chat — Modal or /chat/[matchId]: messages via Supabase Realtime, End chat (both users close → match/messages deleted), 10‑minute expiry, countdown.

---

License

MIT. Built at DeerHacks 2026.