Daily dental health monitoring, powered by agentic AI.

Dental disease is almost entirely preventable, but only if you catch it early. Most people don't find out their habits are putting them at risk until they're already in pain and facing a bill they weren't expecting.

This hits hardest for people without dental insurance. No coverage means no routine cleanings, no early warnings, and small problems that quietly become expensive ones.

DentaGuide is that tool. Under 60 seconds a day, it tracks your habits, symptoms, and weekly photos to catch risk patterns before they develop into something serious.

Every daily check-in fires three AI agents simultaneously:

• Habit Agent — scores brushing, flossing, mouthwash, and sugar intake. Tracks streaks over time.
• Risk Agent — cross-references today's symptoms with 30 days of history and the user's specific tracked conditions. Flags patterns associated with early cavities, gingivitis, or enamel erosion.
• Coach Agent — generates one specific, actionable tip based on what the user is actually tracking. Personalised to their conditions, not a generic reminder to floss.

Users can also upload a weekly photo of their teeth. Gemini AI compares it to last week's and notes any visible changes such as gumline redness, discolouration, buildup as a screening tool, not a diagnosis.

• Node.js 18+
• Python 3.11+
• A Supabase project
• An Gemini API key

git clone https://github.com/your-username/dentaguide.git
cd dentaguide

In your Supabase project, go to SQL Editor and run the following:

-- Profiles table
create table if not exists profiles (
  user_id            uuid primary key references auth.users(id) on delete cascade,
  age                int,
  brushing_frequency text,
  flossing_frequency text,
  last_dentist_visit text,
  previous_treatments text,
  conditions         jsonb default '[]',
  ongoing_issues     text,
  medications        text,
  allergies          text,
  medical_procedures text,
  dental_habits      text,
  onboarding_complete boolean default false,
  created_at         timestamptz default now(),
  updated_at         timestamptz default now()
);

alter table profiles enable row level security;

create policy "Users can manage their own profile"
  on profiles for all
  using (auth.uid() = user_id)
  with check (auth.uid() = user_id);

-- Check-ins table
create table if not exists check_ins (
  id               uuid primary key default gen_random_uuid(),
  user_id          uuid references auth.users(id) on delete cascade,
  brushed          boolean default false,
  flossed          boolean default false,
  mouthwash        boolean default false,
  sugar_intake     text,
  symptoms         text[],
  photo_url        text,
  habit_score      int,
  streak           int,
  risk_severity    text,
  risk_flags       text[],
  risk_explanation text,
  coach_tip        text,
  dental_score     int,
  created_at       timestamptz default now()
);

alter table check_ins enable row level security;

create policy "Users can manage their own check-ins"
  on check_ins for all
  using (auth.uid() = user_id)
  with check (auth.uid() = user_id);

-- Risk flags table
create table if not exists risk_flags (
  id          uuid primary key default gen_random_uuid(),
  user_id     uuid references auth.users(id) on delete cascade,
  severity    text,
  flags       text[],
  explanation text,
  resolved    boolean default false,
  created_at  timestamptz default now()
);

alter table risk_flags enable row level security;

create policy "Users can manage their own risk flags"
  on risk_flags for all
  using (auth.uid() = user_id)
  with check (auth.uid() = user_id);

1. Go to Storage in your Supabase dashboard
2. Click New bucket
3. Name it photos
4. Toggle Public bucket ON
5. Click Create bucket

1. Go to Authentication → Providers → Email
2. Turn off Confirm email
3. Save

From your Supabase project dashboard:

• Project URL → Settings → API → Project URL
• Service role key → Settings → API → service_role (secret key, not the anon key)

cd backend

python3.11 -m venv venv
source venv/bin/activate        # Mac/Linux
# venv\Scripts\activate         # Windows

pip install -r requirements.txt
pip install httpx               # if not already in requirements.txt

Create backend/.env with the following:

SUPABASE_URL=https://your-project-ref.supabase.co
SUPABASE_SERVICE_ROLE_KEY=your-service-role-key
GEMINI_API_KEY=your-gemini-api-key
JWT_SECRET=any-random-string-you-choose

⚠️ Never commit this file. It's already in .gitignore.

uvicorn main:app --reload --port 8000

You should see:

INFO:     Uvicorn running on http://127.0.0.1:8000

Verify it's working: http://localhost:8000 should return {"status": "ok"}.

Open a new terminal tab:

cd frontend

npm install

Create frontend/.env.local:

VITE_API_URL=http://localhost:8000

npm run dev

Open http://localhost:5173 in your browser.

1. Click Get started free and create an account
2. Complete the 3-step onboarding (dental history, conditions, health background)
3. Go to Check-in and log your first daily check-in
4. Optionally upload a photo of your teeth for Gemini visual analysis
5. View your results, dashboard, and trends

dentaguide/
├── frontend/
│   ├── src/
│   │   ├── api/          # API call functions (auth, checkins, profile)
│   │   ├── components/   # Shared UI components (Navbar, HabitRow, Icons)
│   │   ├── context/      # AuthContext (JWT session management)
│   │   ├── pages/        # All page components
│   │   └── index.css     # Global design system (CSS variables)
│   ├── .env.local        # VITE_API_URL (create this)
│   └── package.json
│
└── backend/
    ├── agents/
    │   ├── habit_agent.py    # Scores brushing, flossing, sugar — calculates streak
    │   ├── risk_agent.py     # Gemini AI risk pattern detection
    │   └── coach_agent.py    # Gemini AI personalised tip generation
    ├── routers/
    │   ├── auth.py           # Register, login, logout, update user
    │   ├── profile.py        # Onboarding steps 1/2/3, profile read/update
    │   └── checkins.py       # Today, history, trends endpoints
    ├── photo_agent.py        # Gemini vision — dental photo analysis
    ├── orchestrator.py       # Runs all 3 agents in parallel (asyncio.gather)
    ├── main.py               # FastAPI app, /api/checkin route
    ├── db.py                 # Supabase client
    ├── models.py             # Pydantic request/response models
    ├── config.py             # Loads env vars
    ├── .env                  # Your secrets (create this, never commit)
    └── requirements.txt

• Tharjiha Suthekara — dashboard format, daily check-in flow, photo history analysis, bugfixes, account setting
• Jennifer Huang — dashboard information connection, authentication, profile onboarding, trends page functionality, API integration