Addy Osmani

Who they are

Addy Osmani is an Engineering Leader at Google — ran developer experience for Google Chrome for over a decade, then moved into Google Cloud AI, and now runs the 'Elevate' Substack with 600,000 readers and built agent-skills, a production-grade engineering skills library for AI coding agents.

Person

Addy joined Google in April 2009 — spending the bulk of his tenure as an Engineering Lead on the Google Chrome team, owning developer experience and web performance. He wrote the book on engineering leadership there, literally: 'Leading Effective Engineering Teams' has its own companion site at leet.addy.ie, was featured in the Pragmatic Engineer, and earned him a talk at LeadDev San Francisco in 2022. His move to Google Cloud AI — overseeing Gemini, Vertex, and AI agents — is documented in a blog post titled 'Hello, Gemini.' on his site at addyosmani.com. His public voice is prolific and cross-platform: he keeps a blog at addyosmani.com, writes on Medium, and runs the 'Elevate' Substack at addyo.substack.com with 600,000 readers, covering AI engineering, developer tooling, and engineering leadership. He's built in the open throughout: side projects include agent-skills (production-grade engineering skills for AI coding agents on GitHub), OpenClaw/VisionClaw (open-source AI agent perception), and a Stoic philosophy book titled 'Stoic Mind.' He spoke at the AI Engineer World's Fair on 'Don't build agents you can't answer for,' and at Google I/O 2026 — the through-line across everything is raising the floor on how engineers think about quality, accountability, and high-leverage work.

Company

Google's most recent strategic moment is the Google I/O 2026 announcement, where the company unveiled over 100 new AI-powered features including Gemini 3 models integrated into Search, Gmail, YouTube, Docs, and Chrome, along with AI agents providing real-time, multitasking, and persistent task tracking across Google services. To finance this push, Google raised over $30 billion in debt in May 2026, signaling an aggressive bet to reclaim AI leadership. Google Cloud revenue grew 63% year-over-year in Q1 2026, outpacing both AWS and Azure in quarterly growth rate. In 2025, Google made its largest acquisition ever — buying cloud security company Wiz for $32 billion — and also acquired Intersect for clean-energy data center infrastructure, with that deal closing in 2026. The company committed €5.5 billion to data centers and offices in Germany between 2026 and 2029.

Market

Google holds approximately 90% of the global search market as of March 2026, but faces intense AI rivalry from Microsoft Azure and OpenAI in cloud and search, with Microsoft Bing holding about 3.4% of search. The competitive pressure is compounded by regulatory headwinds: ongoing U.S. antitrust litigation that could force divestiture of Chrome or Android, EU Digital Markets Act compliance costing $430 million annually, and a China antitrust probe launched in response to U.S. tariff action in 2025. Google develops its own TPU chips to insulate against NVIDIA dependency and the U.S.-China chip war, while Microsoft, Amazon, Meta, and Apple are all investing heavily in AI infrastructure.

Network

Network edges aren't available in the data, but Addy's public affiliations are clear: he's a contributor and featured community member at LeadDev, a published author at O'Reilly Media and Smashing Magazine, and has been featured in Gergely Orosz's Pragmatic Engineer newsletter. He is listed as an advisor or community contributor at Vets Who Code.

Sundar Pichai· CEO, Google and Alphabet
Gergely Orosz· Author, The Pragmatic Engineer newsletter

How they likely show up

Long tenure at Google since April 2009, spanning Chrome through Cloud AI — thinks in multi-year investment arcs, not sprints.
Runs the 'Elevate' Substack at 600,000 readers alongside a full-time senior role → high output, high agency; not someone who waits for institutional permission to publish a view.
Public writing themes are consistently practical — high-leverage activities, effective teams, agent accountability — rather than abstract or visionary, suggesting he values concrete frameworks over grand narratives.
Gave a talk titled 'Don't build agents you can't answer for' at the AI Engineer World's Fair → likely skeptical of AI hype that outruns engineering rigor; accountability is a recurring signal.
Moved from Chrome developer experience into Google Cloud AI (Gemini, Vertex, AI agents) → tracks technology transitions closely and positions himself at the intersection of developer tooling and emerging platforms.
Wrote 'Stoic Mind' and lists Stoicism as a public theme → possibly brings a measured, equanimity-oriented lens to leadership pressure and ambiguity.

Conversation tips

→ Reference a specific post or chapter — he's written extensively at addyosmani.com and the Elevate Substack; showing you've read something specific will land far better than generic praise.
→ Ask about the Chrome-to-Cloud-AI transition — the 'Hello, Gemini.' blog post signals he thought carefully about that move and will have a considered story to tell.
→ Lead with engineering accountability or agent design — his AI Engineer World's Fair talk ('Don't build agents you can't answer for') is the clearest statement of what he actually cares about in AI right now.
→ Don't pitch him on 'moving fast' as a virtue — his writing consistently emphasizes high-leverage, deliberate work over velocity for its own sake.
→ If you've built with agent-skills or read the Elevate Substack, mention the specific piece — he'll know immediately whether you actually engaged with the content.

Toolbox

Openers

Open on agent-skills — he built a production-grade engineering skills library for AI coding agents and open-sourced it on GitHub; it's a pointed bet that agent quality needs to be engineered, not hoped for.
Reference his AI Engineer World's Fair keynote, 'Don't build agents you can't answer for' — it's his clearest public statement on AI accountability and a sharp conversation starter about where the industry is getting it wrong.
Mention the '14 More Lessons from 14 Years at Google' post at addyosmani.com — it's a rare long-form retrospective from someone who's actually been inside a large engineering org through multiple technology transitions.

Discovery questions

You've written about high-leverage engineering activities for years — how has your definition of 'high leverage' actually shifted now that AI agents are in the development loop?
The 'don't build agents you can't answer for' framing is striking — what does accountability actually look like in practice when you're shipping agent-powered features at Google's scale?
You moved from Chrome's developer experience into Google Cloud AI overseeing Gemini and Vertex — what surprised you most about the transition from tooling for external developers to building AI products for enterprise customers?

Avoid

Don't lead with generic 'AI will 10x productivity' framing — he is publicly focused on engineering accountability, deliberate practice, and quality standards; breathless AI hype is exactly what his writing pushes back against.

Make it yours