Cookbooks
End-to-end examples for common use cases, with working code in Python, TypeScript, Go, and Ruby. Each cookbook builds a complete workflow you can run locally and adapt to your own project.
These guides assume you have the Hatchet SDK installed and a worker running. If you haven’t done that yet, start with Get Started.
AI
Build agents that reason, act, and observe in a loop. Covers reasoning loops, routing, multi-agent orchestration, and parallelization.
AI AgentsPause a task for human review or approval, freeing the worker slot until the decision arrives. The task resumes exactly where it left off.
Human-in-the-LoopIngest documents, split into chunks, generate embeddings, and write to a vector database using a DAG workflow.
RAG & Data IndexingChain multiple model calls with validation, retries, and structured outputs. Each step is a durable task with independent retry and rate-limit controls.
LLM PipelinesData & Processing
Run the same operation across a large set of items with fan-out, retry, and concurrency control.
Batch ProcessingParse PDFs and files, extract content with OCR or LLMs, validate, enrich, and classify using a DAG pipeline.
Document ProcessingFetch, process, and store web content with retries, timeouts, rate limits, and cron scheduling for recurring refreshes.
Web Scraping