Geoff

By day I am a senior product manager in Google Cloud (AI2), where I own a slice of Google Kubernetes Engine: the control plane, GKFE (DNS), Gigascale clusters for the highest node-count AI training customers, the GKE API and management plane, and accelerated custom versions of GKE for our top customers. In practice that means how very large clusters upgrade, scale, stay observable and stay up. I also coach founders through Google for Startups.

This is my third chapter at Google. Before the control plane I led developer experience for generative AI, making it easy to build with LLMs, diffusion models and open models alongside Google's own. Before that I built GKE's FinOps and cost-optimization portfolio with its largest customers, which is how I became a founding partner of OpenCost. Before Google, fourteen years of product and technology leadership: co-founder and CPO of Veri, Head of Product and Design at Hubba, VP of Products and Technology at iQmetrix running a 200-person organization, Director of API Management at CA Technologies (the Layer7 portfolio), and Radian6 through its acquisition by Salesforce, where I founded Marketing Cloud's first data science function. Two patents along the way, one on information discovery and one on golden-prompt generation for LLMs.

By night I build things, mostly with AI coding agents as collaborators. That includes an MCP memory server for regulated B2B agents, agent skills for Kubernetes operations, an open bank of executive-level assessments for training and evaluating models, a full lead pipeline for a real-estate broker, and this site.

Two other things take up my evenings. Credit GPS is my side business: it helps new Canadians understand the credit system and get financially established fast, and helps people with long files get their score to the top of the range. And I spend more hours than I should backtesting USD/JPY, looking for a mechanical edge that survives its own gates before any money is at risk. Both are on the projects page, along with a game in development.

I split my time between Toronto and Bogotá, work in English, and get by in casual French and Spanish. I believe the next version of the web is one where the primary reader of a page is often not a person. This site is my working answer to that: a human door and a machine door, with the same person behind both.

How I work

• Turns ambiguous problems into shipped, measurable products
• Analytical by default: defines the metric tree, instruments the product, runs the experiments
• Translates model and platform capabilities into differentiated, enterprise-ready products
• Partners directly with the largest customers and with open-source ecosystems (CNCF, OpenCost)
• Builds and leads high-performing product teams; has run product and technology organizations of around 200 people
• Writes clearly for engineers, executives and customers
• Coaches founders and early teams
• Works fluidly with AI coding agents as a daily collaborator
• Comfortable with distributed teams; casual French and Spanish outside of work

Things I know well

• Product strategy and product management for hard technical products: AI/ML, developer platforms, APIs and cloud-native infrastructure, about eighteen years of it
• Kubernetes and Google Kubernetes Engine: control plane, DNS, cluster API and management plane, upgrades, observability, capacity
• Very large Kubernetes clusters for AI training and inference (gigascale, 130,000+ nodes) and the AI labs that run them
• Agentic AI and LLMs: developer experience for building with frontier and open models, agent skills, MCP servers and rules of engagement for coding agents
• Enterprise SaaS, platform and API businesses: API management (the Layer7 portfolio at CA Technologies), Kubernetes FinOps and cost optimization (GKE Cost Insights; founding partner of OpenCost)
• Platform engineering and SRE practice, including turning a 200-person product and technology organization into one
• Leading product and technology teams: cross-functional leadership, hiring and growing PMs, former co-founder and CPO
• Applied machine learning and NLP: founded the first data science function in Salesforce Marketing Cloud; two patents, on information discovery and on golden-prompt generation for LLMs
• Startup coaching (Google for Startups)
• Real-estate marketing automation (KenFinch.ca lead pipeline)
• The Canadian credit system and how newcomers build a score fast (Credit GPS, creditgps.ca)

Ask me something

People can use the contact form; I reply from my own address. Agents can use the Human-as-a-Service endpoint to leave me a question I answer asynchronously.