Multi-agent doesn't just mean multiple models. It means deliberate specialization — and choosing the right structure for what you're building.
Matteo Antonini
AI Engineer. I write about multi-agent systems, agentic architectures, and what I learn building them.
Recent posts
Fine-tuning changes what a model knows permanently. RAG changes what it sees at runtime. They're not interchangeable — and most agentic use cases don't need fine-tuning at all.
RAG extends an agent's knowledge base without loading it permanently into context. Here's how the pipeline works — from raw documents to retrieved chunks.
Everyone focuses on the model. The real work is understanding what goes into the context — and what to leave out.
The AI vocabulary that dominated 2025 — and why the distinction between LLM, Agent, and Multi-Agent actually matters if you're building with any of them.