Writing
Articles
Longer pieces on AI agents that have to run in production, written from client work and from talks. Newest first.
I'm Matteo Gazzurelli. Through MaGa Srl, in Brescia, I work as fractional CTO and agentic AI consultant, and I publish here what I otherwise explain twice: once in a meeting, once on stage.
The Company Brain for AI Agents: Beyond One Person
What changes in permissions, decisions and deletion when people and AI agents share knowledge across departments, and where things stand in October 2026.
From CrewAI to Production Agents
What a role-based multi-agent framework gave me at prototype stage, what production asked for on top, and how that work moved to Google ADK and the Claude Agent SDK.
Google ADK in 2026: Where It Fits in Gemini Enterprise Agent Platform
Google kept the ADK name. The four names in Google's agent stack, what each one does, and which question each one answers.
The Developer's Role in the Age of AI: From Executor to Director
In 2021 you wrote code line by line. In 2026 you write the spec and direct the agents that run it. What that changes in a developer's day, and what stays.
Guides
Two reference pages I keep dated and update when the platforms change.
Google vs Anthropic for AI agents
Managed platform, framework, models, hosting, access and integrations, layer by layer. Facts first, my view marked as such.
Google ADK integration: evaluate, integrate, operate
When ADK makes sense, how it connects to the systems you already have, and what running it after launch asks of your team.
Also on Medium
Articles I published on Medium, including the pieces written for the Google Cloud publication. The links open on medium.com.
