Player not loading? Watch on YouTube
Kayla from Jotform compares five tools for coordinating agents, retrieval and workflow execution. She explains the differences between agent, RAG, workflow and cloud orchestration before describing each tool's approach.
LangGraph uses graph-based execution and shared state for branching, long-running workflows. The presenter describes checkpointing for inspecting previous steps, while noting that its structure can feel unfamiliar to users of simpler pipelines. CrewAI organizes work around agents with defined roles, goals and tasks. She presents it as useful for prototyping, with less detailed control flow and an additional routing feature.
LlamaIndex focuses on preparing and retrieving data for RAG, including document splitting, metadata, ranking and filtering. The comparison notes less control over broader orchestration. Haystack connects components such as retrievers and prompt builders into pipelines that can be saved and versioned, but requires more setup decisions.
Prefect adds tasks and flows to Python functions through decorators, with retries, logging and error handling. The speaker says workflows can run on local machines, containers or cloud infrastructure. This is an orchestration overview rather than a local LLM setup guide. The closing section promotes Jotform forms and AI Agents as ways to collect structured inputs and trigger workflows through integrations and webhooks.
