A banking intelligence platform that analyzes customer transaction data to detect life transitions and equips relationship managers with AI-generated outreach briefs.
LifePulse monitors customer spending patterns to identify significant life events — relocations, new babies, home purchases, job changes, financial stress, marriages, and divorces. When a transition is detected, relationship managers receive a personalized outreach brief with recommended products, talking points, and a draft message.
Built as a Capital One internship interview demo.
LifePulse/
├── backend/ # Python FastAPI — data generation, detection, brief API
└── frontend/ # React + TypeScript (Vite) — dashboard UI
- Synthetic transaction generator — 12 seeded customers with 90 days of realistic transaction history and embedded life-event signals
- Rule-based event detector — flags customers based on spending pattern rules (merchant categories, frequency, amount thresholds)
- Outreach brief generator — produces personalized briefs with event summary, transaction signals, recommended products, and draft messaging
- Dashboard UI — sortable customer list with confidence scores, severity badges, and one-click brief viewer
cd backend
python -m venv venv
source venv/bin/activate # Windows: venv\Scripts\activate
pip install -r requirements.txt
uvicorn main:app --reload --port 8000API available at http://localhost:8000 — Swagger docs at /docs.
cd frontend
npm install
npm run devApp available at http://localhost:5173.
| Method | Endpoint | Description |
|---|---|---|
| GET | /api/customers |
All customers with detected life events |
| GET | /api/customers/{id} |
Single customer detail |
| GET | /api/customers/{id}/transactions |
Full transaction history |
| GET | /api/customers/{id}/brief |
AI-generated outreach brief |
| Event | Key Signals |
|---|---|
| Relocation | Moving truck rental, storage, new utility setup, furniture stores |
| New Baby | Baby retailers, OB/GYN, prenatal care, pediatric visits |
| Home Purchase | Title insurance, escrow, home inspection, Home Depot spike |
| Job Change | Deposit source change, LinkedIn Premium, professional apparel |
| Financial Stress | NSF/overdraft fees, payday loans, late fees, spending decline |
| Marriage | Wedding vendors, jeweler, honeymoon travel |
| Divorce | Legal fees, mediation, new apartment deposit |
- Backend: Python 3.11, FastAPI, Pydantic v2
- Frontend: React 18, TypeScript, Vite, Tailwind CSS
- Auth: Mock (hardcoded session — no real auth for demo)