Hey everyone 👋 Tarun here, maker of Orchflow. I kept running into the same problem while building AI apps — every time I needed multi-step workflows, I ended up rebuilding the same orchestration layer: * handling dependencies * managing state * adding retries * wiring multiple agents together It was slow, messy, and honestly… frustrating. So I built **Orchflow**. It’s an open-source API that lets you run AI workflows as real DAGs — not just chains. You can define tasks, dependencies, human approval steps, and let Orchflow handle execution, retries, and routing. What makes it different: • Real DAG-based execution (not linear flows) • Multi-agent orchestration • Human-in-the-loop checkpoints • Redis + Postgres backed state (fast + durable) • Simple REST APIs to integrate anywhere The goal is simple: 👉 Stop rebuilding workflow infrastructure. Just focus on your AI logic. Would love your feedback or questions 🙏

Orchflow

Open-source API for AI workflow orchestration

Open-source API for AI workflow orchestration