LiveFull Stack2023
Pulse Analytics
Real-time analytics platform processing 2M+ events/day. Distributed architecture on AWS cut reporting latency from 8 s to 180 ms.
About this project
Pulse Analytics is a multi-tenant SaaS platform that ingests, processes, and visualises high-volume event streams in real time. The ingestion layer is built on Kafka with a custom schema registry; the query engine uses materialised views in PostgreSQL backed by TimescaleDB for time-series acceleration. The frontend is a Next.js App Router application with live charts powered by a WebSocket feed, streamed directly from an Edge Function.
Technical highlights
Kafka consumer group auto-scaling based on partition lag
Row-level security so each tenant only sees their data
P95 query time under 200 ms even at peak load
Zero-downtime schema migrations with shadow tables
More projects