Blog
Longer pieces on what actually holds up in production — loop engineering, spec-driven development, multi-agent tradeoffs, and the parts of the job that moved rather than disappeared. Newest first.
Google did not rename ADK. See how the code-first framework fits into Gemini Enterprise Agent Platform, Agent Runtime, Cloud Run and Gemini Enterprise.
Read the article →
In 2021 you wrote code line by line. In 2026 you write the spec and direct the agents that execute it. What actually changes for a working developer — and what doesn't.
Read the article →
Articles published on Medium, including pieces written for the Google Cloud publication. These links open on medium.com.
The shift from implementor to conductor, and why the harness around the model matters more than the model itself. TL;DR AI compresses implementation, not judgment: requirements…
Read on Medium
If AI writes all the code, why does Anthropic still hire developers? Why I wrote this Because I’m tired. Tired of watching smart, capable people — founders, managers…
Read on Medium
A field guide for anyone who has been building agents on Google Cloud and just watched the platform get a new name — and a new shape. At Cloud Next ’26 , Google renamed Vertex AI…
Read on Medium
Google ADK: From 0 to Agentic AI — Part 12 TL;DR Building an agent that works locally is different from deploying one that customers depend on. This final article covers the…
Read on Medium
Google ADK: From 0 to Agentic AI — Part 11 TL;DR Building agents by hand doesn’t scale. In Part 11, we climb from “tutorial” to “architecture” — implementing a system where users…
Read on Medium
Google ADK: From 0 to Agentic AI — Part 10 TL;DR Agents that run in production need frontends that handle streaming text, tool calls mid-response, and state updates — all in…
Read on Medium
Google ADK: From 0 to Agentic AI — Part 9 TL;DR Chat is the default interaction mode, but it’s insufficient for most real applications. Generative UI (A2UI) enables agents to…
Read on Medium
Google ADK: From 0 to Agentic AI — Part 8 TL;DR While MCP enables agents to call tools, the Agent-to-Agent (A2A) Protocol enables agents to delegate work to other agents across…
Read on Medium
Google ADK: From 0 to Agentic AI — Part 7 TL;DR The Model Context Protocol (MCP) functions as a universal standard for connecting agents to tools and external systems. Rather than…
Read on Medium
Google ADK: From 0 to Agentic AI — Part 6 TL;DR Part 5 gave agents memory (they remember conversations). This article gives them a library: the company knowledge base via RAG…
Read on Medium