GitHub, live
Connecting

Powered by

What’s happening
on GitHub?

Explore GitHub activity, pull requests, and AI insights.

PushPull requestStarCommentOther
—

Captured · last 24 hours

—events · 10s
—pull requests · 10s
—stars · 10s
—comments · 10s
—lines merged · 10s

Waiting for captured ten-second counts.

Landing now

Loading
Rolling 24h feed · merged PRs as they land—

Waiting for captured PRs

A little closer to the action.

Waiting for results

Activity · 1 hour

The wider picture.

Loading

A rolling 24 hours of captured events, grouped by their original UTC hour. Edge hours cover only part of an hour.

—captured events

Rolling 24-hour result

Loading…

Events, hour by hour

UTC

What the stream is made of

Event types in the rolling 24-hour result.

Inside the pull requests.

Merged pull requests

Languages in the mix

Repository languages

Language results are not available yet.

How big is a change?

Average lines changed

—lines changed

From open to merged.

Average time to merge

—among pull requests with timing data

Pipeline snapshot

Data, in motion.

Total collected
— GB

Compressed source data collected to date

— GB before gzip compression

— · — compression

Avg. ingestion delay
— sec

Collector → decode · recent merged PRs

Daily collection
— GB/day

Previous completed UTC day

— GB before gzip compression

Waiting for pipeline snapshot

How it gets here.

Public GitHub events, streamed into an open Iceberg table by Eon Stream and queried live by Eon Query, both part of Eon Data Engine.

01

GitHub

The public events feed, read continuously. Merged PRs and sampled pushes are fetched for their title, description and size.

02

Object storage

Events land in S3 in small batches. Each new file sends a notification.

03

Eon Stream

  • Reads each file as soon as its notification arrives.
  • Enriches rows in flight: AI-attribution rules on every merged PR, and a language model reads a sample of PRs and pushes.
  • Keeps live aggregates up to date as rows land.
  • Commits every second, exactly once.
04

Iceberg table

An open table on S3 in the AWS Glue catalog. Any engine can read it.

05

Eon Query

SQL over the table every few seconds, straight from the lake.

06

This page

Results stream to your browser as they change.