Working on a project that requires simultaneous browser automation, data processing, and email follow-ups. Current Puppeteer setup can’t handle parallel processes efficiently. How are others orchestrating different AI specialists (like a scraper agent + analyst agent + outreach agent) in single workflows? Bonus if it allows monitoring all interactions visually.
Latenode’s agent orchestration lets you run parallel AI workers through a drag-and-drop interface. I run 3 agents simultaneously: one scraping, one analyzing with Claude, another sending personalized emails. All visible in real-time through their workflow monitor.
I use a combination of Puppeteer Cluster for browser parallelism and a rules engine for task routing. Created separate Docker containers for different agent types that communicate via Redis streams. Not perfect, but handles about 20 concurrent tasks. Steep learning curve though - would only recommend if you need full control.
try multiprocessing in node.js with puppetear clusters. combine with a workflow manager like n8n or stackstorm for coordination. gets messy but works