Instant segments. Live journeys. Anomalies the moment they happen.
Web-Pop ingests your raw web events and answers the questions you already ask — who buys, where they drop off, what looks fraudulent, what looks automated — without a pipeline, without a warehouse, without waiting for a batch job.
Buyers, high-intent users, engaged users, bounce traffic, mobile, organic — ready in milliseconds, not after a nightly job.
Page view → add to cart → purchase, in order, per session. Where your visitors drop off, and what converts.
Purchases that are abnormally high for their own traffic source. Flagged the moment they land, with a confidence score.
Visitors whose page-view timing is impossibly fast. Separated from real humans without a single rule.
The most common event sequences — and the rarest, which are usually the interesting ones.
When your traffic and revenue actually happen, hour by hour.
The core technology behind Web-Pop is under patent protection, so we can't share how it works — only what it returns. Every endpoint exposes human-readable insights: counts, names, percentages, confidence scores. The internals stay ours.
Read-only, JSON in and out. No write operations, no configuration — the demo is a fixed snapshot of one dataset.
| Method | Endpoint | Returns |
|---|---|---|
| GET | /summary | Events, users, sessions, purchases, buyers, bounce rate, average order value |
| GET | /segments | Named audience segments with user counts |
| GET | /funnel | Purchase journey steps: users, conversion %, drop-off % |
| GET | /anomalies?z=3&limit=20 | Revenue outliers with opaque ids and confidence scores |
| GET | /bots?z=3&limit=20 | Automated visitors with opaque ids and timing stats |
| GET | /patterns?n=3&k=10 | Most frequent and rarest event sequences |
| GET | /peak-hours | Event volume by hour of day |
| GET | /health | Service status and uptime |
Feedback