Parallel AI coding agents with live previews
AI coding assistants are powerful but sequential - you can only run one at a time. For large projects, this creates a bottleneck. Developers needed a way to parallelize AI-assisted development while maintaining visibility into what each agent is doing.
Built a real-time orchestration platform using Elixir/Phoenix and LiveView. Each agent runs in its own supervised process, with live terminal output streamed to the browser via WebSockets. Phoenix's built-in PubSub handles the real-time updates with zero additional infrastructure.
Start with Estimo and get a first-pass estimate without the usual back and forth.
Get an estimate →