Spendly – WhatsApp-Powered Expense Tracker

A WhatsApp-powered expense tracker using AI to parse natural language messages into structured spending data with a real-time analytics dashboard.

Overview

Spendly is a microservices-based expense tracking system that lets users log expenses via WhatsApp messages. AI parses natural language into structured spending data, and a real-time analytics dashboard visualizes spending patterns. The architecture uses RabbitMQ for async messaging, Hookdeck for webhook ingestion, and Docker Compose for production-grade orchestration.

Key Achievements

Built the expense logging microservice using Node.js, RabbitMQ, and LangChain.js with Hookdeck for webhook ingestion and rate-limited AI processing.

Developed analytics and auth microservice using FastAPI with email auth, WhatsApp link verification, and Google OAuth.

Built the frontend dashboard using Next.js 15 and Recharts with interactive spending visualizations, WhatsApp OTP-based phone linking, and cookie-based auth with email verification.

Wrote complete Infrastructure as Code using Docker Compose with multi-service orchestration for production-grade deployment.