Automatically matches businesses to class action settlement claims they're eligible for. Connect your accounting software, and we find money you're owed.
- Sign up and connect your QuickBooks/Xero/NetSuite via OAuth (2 min)
- Background engine continuously scans open settlements and matches against your vendor purchase history
- Email notification when a match is found — one-click approve or decline
- We file the claim automatically via claims administrator portals
- You get paid — we charge 25% only when you receive money
CourtListener API (daily poll)
+
TopClassActions scraper (daily)
+
ClassAction.org scraper (daily)
+
Claims Admin portals scraper (daily)
↓
Settlement DB (Postgres + Prisma)
↓
Rutter API → vendor transactions from accounting software
↓
Matching engine (exact → fuzzy → LLM entity resolution)
↓
Email approval flow (signed JWT links, no login required)
↓
Playwright form automation → file with claims admin
↓
Stripe → auto-charge fee when payout confirmed
| Layer | Tool |
|---|---|
| Framework | Next.js 14 (App Router) |
| Auth | Clerk |
| Database | PostgreSQL + Prisma |
| Accounting Integration | Rutter API |
| Resend + React Email | |
| Background Jobs | Inngest |
| Matching | Fuse.js (fuzzy) + OpenAI (entity resolution) |
| Form Automation | Playwright |
| Payments | Stripe |
| Styling | Tailwind CSS |
src/
├── app/ # Next.js App Router
│ ├── api/
│ │ ├── claims/respond/ # Email approval link handler
│ │ ├── company/ # Company CRUD
│ │ ├── inngest/ # Inngest webhook endpoint
│ │ ├── rutter/ # Rutter OAuth flow
│ │ └── webhooks/stripe/ # Stripe webhooks
│ ├── dashboard/ # Claims dashboard (read-only ledger)
│ ├── onboarding/ # Company setup + accounting connection
│ ├── sign-in/ # Clerk sign-in
│ └── sign-up/ # Clerk sign-up
├── lib/
│ ├── approval/ # JWT token generation + claim response flow
│ ├── email/
│ │ ├── send.ts # Email sending via Resend
│ │ └── templates/ # React Email templates
│ ├── filing/
│ │ ├── base.ts # Base Playwright filer
│ │ ├── admins/ # Per-admin Playwright scripts
│ │ └── index.ts # Filer registry
│ ├── ingestion/ # Settlement data scrapers
│ │ ├── court-listener.ts # CourtListener API
│ │ ├── top-class-actions.ts # TopClassActions.com scraper
│ │ ├── class-action-org.ts # ClassAction.org scraper
│ │ └── claims-admin-scrapers.ts # Epiq, Rust, JND, Kroll, Angeion
│ ├── inngest/ # Background job definitions
│ ├── matching/
│ │ ├── engine.ts # Full matching pipeline
│ │ ├── fuzzy.ts # Fuse.js fuzzy matching
│ │ └── llm-resolver.ts # OpenAI entity resolution
│ ├── rutter/ # Accounting platform integration
│ │ ├── client.ts # Rutter API client
│ │ └── sync.ts # Transaction sync + normalization
│ ├── stripe/ # Payment / fee collection
│ ├── db.ts # Prisma client singleton
│ └── env.ts # Environment validation
├── middleware.ts # Clerk auth middleware
prisma/
└── schema.prisma # Database schema
-
Copy environment variables:
cp .env.example .env
-
Fill in all API keys in
.env -
Install dependencies:
npm install
-
Set up the database:
npx prisma db push
-
Run the dev server:
npm run dev
-
In a separate terminal, start Inngest:
npm run inngest:dev
- Email as the UI: The approval email is the most important UI surface. Users rarely need to log in.
- Signed JWT links: No login required to approve/decline claims. Token in the URL is the auth.
- Rutter over individual integrations: One API covers QuickBooks, Xero, NetSuite, Sage, etc.
- LLM for entity resolution: Fuzzy string matching misses subsidiaries and DBA names. A small LLM call handles "is Hillshire Brands the same as Tyson Foods?" cheaply.
- Per-admin Playwright scripts: Claims admins don't have APIs. Finite set (~10 major admins) makes per-admin automation feasible.
- Fee on payout only: Stripe charges saved payment method when claim pays out. Zero upfront cost to user.