Realtime Analytics Dashboard cover
All work

2024

Realtime Analytics Dashboard

A streaming metrics dashboard that ingests events and renders live charts under a second of latency.

Overview

Built an end-to-end analytics pipeline that captures product events, aggregates them in a columnar store, and pushes updates to a browser dashboard over websockets. Designed the schema, the ingestion worker, and the frontend visualisation layer. Focused on keeping the render path fast even with tens of thousands of events per minute.

Role & impact

My role

Full-stack engineer — owned data model, ingestion, and UI.

Impact

Cut dashboard load time from 4.2s to 0.6s and enabled the team to spot anomalies in real time.

Stack

TypeScriptReactNode.jsClickHouseWebSocket