Agent SDKs & Frameworks · TRACK 04/05
Orchestration Frameworks
LlamaIndex, CrewAI, AutoGen, Semantic Kernel, Haystack — one guide each.
// THE TRACK
01 · START HERELlamaIndexUnderstand LlamaIndex's pipeline from raw documents to answered questions — loaders, indexes, retrievers, and query engines — in plain terms.BEGINNERAutoGenUnderstand AutoGen's core idea — agents that solve problems by talking to each other — and when conversation beats rigid orchestration.BEGINNERCrewAI Multi-Agent TeamsLearn CrewAI's company-like model — agents with roles and goals working through tasks — and assemble your first working crew.BEGINNERLlamaIndex vs LangChainCompare the two biggest Python LLM frameworks head-to-head on RAG — and learn when the right answer is actually using both together.INTERMEDIATEHaystackUnderstand Haystack's explicit-pipeline philosophy — every component visible and swappable — and the production cases where that pays off.INTERMEDIATESemantic KernelUnderstand Semantic Kernel's model for orchestrating LLMs — plugins, functions, planners, and memory — and why it's the go-to for enterprise .NET and Java AI apps.INTERMEDIATECrewAI vs AutoGenCompare CrewAI's structured role-based crews against AutoGen's free-flowing agent conversations so you can pick the multi-agent model that fits your task's need for control versus flexibility.INTERMEDIATEsmolagentsLearn how Hugging Face's smolagents takes a minimal, code-first approach — the model writes and runs Python to act instead of emitting JSON tool calls — and where that pattern wins.INTERMEDIATE