Real-time Page Analytics Built for Scale

Problem

Qwilr needed customer engagement analytics — how prospects interacted with pages, what they viewed, when they dropped off. The existing approach couldn’t handle the volume or provide real-time insights. The system needed to scale to tens of millions of events per day while keeping queries fast for end-user dashboards.

Action

I led a team of engineers to build a backend analytics system on Kafka and Druid. Kafka handled high-throughput event ingestion, while Druid provided the fast analytical queries needed for real-time dashboards. We built a custom API layer for data access and a dashboard interface so customers could see engagement metrics as they happened.

Outcome

The system handled thousands of requests per second and scaled to tens of millions of events per day. Customers got real-time dashboards showing exactly how their prospects engaged with content — turning Qwilr pages from static documents into measurable sales tools.