Session replay rasterizer

Temporal worker that converts session replays into MP4 videos using Puppeteer + Chromium. Runs on a dedicated rasterization-task-queue.

Architecture

text

Temporal workflow (Python)
  └─ dispatches activity to rasterization-task-queue
       └─ Node.js worker picks it up
            ├─ rasterizeRecording (Puppeteer records the replay to raw MP4)
            ├─ postProcessToMp4 (ffmpeg: strip pre-roll, trim, speed correct)
            ├─ uploadToS3
            └─ return metadata

The worker maintains a warm Chromium browser pool — a single browser instance is reused across exports and recycled after N uses.

Modules

File	Purpose
`index.ts`	Worker entry point — launches browser pool, connects to Temporal
`activities.ts`	Temporal activity: record → postprocess → upload → cleanup
`recorder.ts`	Puppeteer recording logic (puppeteer-screen-recorder)
`postprocess.ts`	ffmpeg post-processing (trim, speed correction, fps)
`browser-pool.ts`	Warm Chromium lifecycle manager
`storage.ts`	S3 upload
`config.ts`	Environment variable configuration
`types.ts`	Input/output contracts

Two pipelines

User video exports: full pipeline with ffmpeg post-processing (strip pre-roll, trim to duration, apply speed correction)
AI/Gemini pipeline: skip_postprocessing: true — uploads raw recording directly

Running locally

bash

bin/temporal-recording-rasterizer-worker

Requires Chromium and ffmpeg installed locally.

Docker

bash

docker build -f Dockerfile.recording-rasterizer -t posthog-recording-rasterizer .

Uses Debian chromium package (not Puppeteer's bundled download).