Untangle is a smart consultation and booking system for textured-hair stylists. Clients complete a structured hair intake form before their appointment. Stylists receive a full profile — hair type, density, porosity, treatment history, preparation status — along with an estimated service time and suggested price range, before anyone sits in the chair.
- Clients underestimating their hair density and booking short slots for long styles
- Arriving unprepared with unwashed, matted hair
- Stylists having to guess the price mid-appointment
- Surprise chemical damage discovered at the chair
Untangle moves the consultation out of the chair and into a structured digital form.
- Stylist creates an account and sets up their services with base prices and estimated hours
- Stylist shares their intake link —
untangle.app/intake/your-name— with a client before the appointment - Client fills out a 5-step intake form covering hair details, history, preparation status, and style goals
- Untangle estimates service time, prep time, and suggested price range using rule-based logic
- Stylist reviews the client profile on their dashboard before confirming the appointment
| Layer | Technology |
|---|---|
| Frontend | Next.js 14 (App Router) |
| Styling | Tailwind CSS with custom petal color palette |
| Icons | Lucide React |
| HTTP client | Axios |
| Backend | FastAPI (Python) on localhost:8000 |
| Database | SQLite (dev) / PostgreSQL (prod) via SQLAlchemy |
| Auth | JWT (python-jose) + bcrypt (passlib) |
| Estimation | Rule-based logic — no ML required |
- Node.js 18+
- Python 3.11+
npm install
npm run devOpens at http://localhost:3000.
cd backend
bash start.shRuns at http://localhost:8000. API docs at http://localhost:8000/docs.
The backend creates a untangle.db SQLite file automatically on first run. No setup required.
.env.local (frontend):
NEXT_PUBLIC_API_URL=http://localhost:8000
Backend environment variables (all optional):
DATABASE_URL=postgresql://user:password@localhost/untangle # defaults to sqlite:///./untangle.db
SECRET_KEY=your-secret-key-here # defaults to a dev key
RESEND_API_KEY=re_xxxxxxxx # if unset, emails print to the console
EMAIL_FROM=Untangle <hello@yourdomain.com> # must be a Resend-verified domain
FRONTEND_URL=http://localhost:3000 # used in email links
| Command | Description |
|---|---|
npm run dev |
Start Next.js dev server |
npm run build |
Build production bundle |
npm start |
Run production server |
untangle/
├── app/ # Next.js App Router pages
│ ├── page.js # Landing page
│ ├── login/page.js # Login
│ ├── signup/page.js # Stylist signup with slug selection
│ ├── onboarding/page.js # Service setup after signup
│ ├── dashboard/
│ │ ├── page.js # Intake list + summary stats
│ │ ├── services/page.js # Manage services (CRUD)
│ │ └── intake/[token]/page.js # Single intake detail + decision actions
│ └── intake/
│ └── [slug]/
│ ├── page.js # Client: choose service, enter details, pick appointment time
│ └── [token]/
│ ├── page.js # Client: 5-step hair intake wizard
│ └── done/page.js # Client: confirmation + estimate
├── lib/
│ └── api.js # Axios client with JWT interceptor
├── backend/
│ ├── main.py # FastAPI routes + lifespan startup
│ ├── models.py # SQLAlchemy ORM models
│ ├── schemas.py # Pydantic request/response schemas
│ ├── database.py # DB connection + session factory
│ ├── auth.py # JWT creation + verification
│ ├── estimator.py # Complexity estimation logic
│ ├── email_service.py # Resend wrapper with console-log fallback
│ ├── email_templates.py # HTML/text templates for all 4 emails
│ ├── scheduler.py # APScheduler jobs (24h reminder, 48h follow-up)
│ ├── tests/ # pytest suite
│ └── requirements.txt
└── tailwind.config.js
| Step | Fields |
|---|---|
| Hair details | Length, density, porosity, strand thickness, condition |
| Hair history | Last relaxer, last color, last heat treatment, breakage |
| Preparation | Washed, detangled, product-free (day of appointment) |
| Goals | Style inspiration link, preferred duration, scalp issues |
| Review | Full summary before submit |
All estimation is rule-based in backend/estimator.py. No machine learning.
estimated_hours = (base_service_hours
× length_mult × density_mult × thickness_mult
+ condition_extra + porosity_extra + relaxer_extra + color_extra + breakage_extra)
rounded to the nearest 0.5h
prep_time = +20m if not washed, +30m if not detangled, +10m if product-heavy
price_range = base_price × the same multipliers, plus flat extras, then ±10%
complexity_score = weighted sum of all factors, capped at 10
Multipliers:
| Factor | Values |
|---|---|
| Length (shoulder = 1.0) | 0.7 (TWA) → 1.6 (waist+) |
| Density | 0.85 (low) / 1.0 (medium) / 1.3 (high) |
| Thickness | 0.9 (fine) / 1.0 (medium) / 1.2 (coarse) |
Flat hour extras:
| Factor | Value |
|---|---|
| Condition | 0h (healthy) · 0.5h (dry) · 0.75h (transitioning) · 1.0h (damaged) |
| Porosity (low) | +0.25h |
| Last relaxer < 6 months / 6–12 months | +0.5h / +0.25h |
| Last color < 3 months / 3–6 months | +0.5h / +0.25h |
| Has breakage | +0.25h |
Untangle sends four transactional emails through Resend. Two fire synchronously via FastAPI BackgroundTasks; two are time-based and dispatched by APScheduler running in-process.
| # | Trigger | Goes to | Mechanism |
|---|---|---|---|
| 1 | Client submits intake | Stylist | BackgroundTasks on POST /intake/{token}/submit |
| 2 | Stylist updates decision (confirm / reschedule / adjust price / request prep) | Client | BackgroundTasks on PATCH /dashboard/intakes/{token}/decision |
| 3 | 24 hours before the appointment | Client | APScheduler every 15 min; idempotent via reminder_sent_at |
| 4 | 48 hours after a pending intake with no stylist action | Stylist | APScheduler every 1 hour; idempotent via followup_sent_at |
Templates live in backend/email_templates.py. Without RESEND_API_KEY set, sends are logged to the terminal instead — useful for iterating on templates locally without burning quota.
The 24-hour reminder reads the prep checklist (is_washed, is_detangled, is_product_free) from the submitted hair profile and only nags the client about steps they actually flagged as incomplete.
| Method | Path | Auth | Description |
|---|---|---|---|
POST |
/auth/register |
— | Create stylist account |
POST |
/auth/login |
— | Get JWT |
GET |
/stylist/{slug} |
— | Public profile + service list |
GET |
/stylist/me/profile |
Stylist | Read own profile |
PUT |
/stylist/me/profile |
Stylist | Update bio, location, instagram |
GET |
/services |
Stylist | List own services |
POST |
/services |
Stylist | Add a service |
PUT |
/services/{id} |
Stylist | Edit a service |
DELETE |
/services/{id} |
Stylist | Remove a service |
POST |
/intake/{slug}/start |
— | Create intake session (captures appointment time), return token |
GET |
/intake/{token} |
— | Get session info |
POST |
/intake/{token}/submit |
— | Submit hair profile, run estimation, email stylist |
GET |
/dashboard/intakes |
Stylist | All submissions |
GET |
/dashboard/intakes/{token} |
Stylist | Single submission detail |
PATCH |
/dashboard/intakes/{token}/decision |
Stylist | Confirm / reschedule / adjust price / request prep, email client |
Backend tests use pytest with an in-memory SQLite per test (via StaticPool so all sessions see the same DB). email_service.send_email is auto-mocked across the whole suite so no test can accidentally hit Resend.
cd backend
./venv/bin/python -m pytest -qCurrent coverage:
| File | What it locks down |
|---|---|
tests/test_estimator.py |
Length, density, thickness multipliers; prep-time accumulation; complexity score cap; defensive defaults |
tests/test_scheduler.py |
Idempotency of both scheduled jobs; window/precondition skip behavior |
tests/test_decision_auth.py |
Stylist A can update their own intake; Stylist B gets a 404 and the data is actually unchanged |
MIT
