Hatchet

Hatchet is an orchestration engine designed for modern applications that require high reliability for background tasks, AI agents, and long-running business processes. It provides ultra-low latency, high-throughput scheduling with built-in concurrency control, fair queuing policies (e.g., FIFO, LIFO, round-robin, priority), and rate limiting to protect downstream systems during traffic spikes.

Hatchet offers resilient execution capabilities: customizable retry strategies, error handling, step-level replay, and durable execution ensure quick recovery from transient failures. It also includes comprehensive observability, with fully searchable run logs, latency and error-rate monitoring, and incremental streaming updates, enabling DevOps teams to track workflow states in real time.

As an open-source project, Hatchet can be self-hosted, and a cloud service (Hatchet Cloud) is available to reduce operational overhead, although publicly documented pricing and scalability guarantees remain limited. Users must weigh the tradeoffs between self-hosting and cloud, and account for the learning curve associated with its workflow DSL.