Paige Bailey is Head of Developer Relations at Google DeepMind. She has a long tenure at Google. Her job sits between the people who build Gemini and the people who have to work out what to do with it.
Her through-line is open source. She contributed to TensorFlow before her first job interview, so the work itself became her interview. It fits a habit of learning by building that goes back to a rescued Apple II, where she typed in text adventures from Byte magazine, line by line.
Paige joined AJ Green and Catherine McMillan on The AI Collective Podcast, and she is exactly the kind of person we built the Hall of Fame for.

From a rescued Apple II to Gemini
Paige trained as a geophysicist. When she told her professors she wanted to write software libraries for earth science, the response was that it was not real science and a waste of her time. She wrote them anyway.
It was the right bet. Geophysics adopted GPUs years before most of the industry, for velocity modeling and fluid dynamics simulations. When TensorFlow first shipped, it ran on CPUs only. Paige already knew CUDA, so she had something useful to offer the team.
From there, she worked across the major orgs everyone knows: Google, Microsoft and GitHub, and Google DeepMind. She spent a year at Microsoft as a director of machine learning and MLOps at GitHub, working on Codespaces, VS Code, and Copilot. Back at Google, she was a lead product manager for generative models, including PaLM 2 and the first Gemini models. In 2024, she moved into developer relations.
Each stop supplied something the next one needed. Physics gave her the accelerators. Open source gave her the habit of building in public. Product work put her next to the model teams. Developer relations is a bit of everything, so she became a generalist across every team she worked with.

The front door to Gemini
Paige now leads developer experience for a family of tools that includes the Gemini API, Gemma and AI Studio. Her job is bridging the gap between where models are today and how people are actually using them, and carrying what those people find back to the teams who build the models.
She treats the second half as the whole point. In her words, “If you don’t close the loop with the product team, then you’re not really doing developer relations.” Her team sits alongside the product group at DeepMind, turning bug reports, forum posts and early-access feedback into fixes and evals.
The audience has changed, too. Developer relations used to mean millions of software engineers. Paige sees it now reaching scientists, teachers, artists, historians and information workers, plus a growing population of agents that need tools they can discover on their own. She calls it a hard distributed systems problem, and a human one: part of the job is behavioral change management, helping people who have worked one way for decades start building something new.

Whimsy as a method
If you want to understand how Paige works, look at her side projects. She has installed solar-powered silent-film birdhouses around San Francisco as street art. She built folklore.dev, which maps oral histories from the Computer History Museum so you can search the moments someone appears on screen or speaks. She made a Wikipedia Racer that generates rabbit holes on demand, and a bot that emails her a roast of her taste in music.
Paige has said she uses her side projects to test new Gemini API features early. The point is as much cultural as technical: with the cost of writing software falling toward zero, she argues taste is the skill that matters most.
Make the benchmark you wish existed
In our conversation, Paige shared her love of benchmarks, and there was one in particular she was adamant about. Food Truck Bench tests whether a model can run a food truck over time. She wants a laundromat version next, for the small business owners deciding whether to open one.
Her advice for anyone who wants models to do something better is to “create an eval for it and then kind of like name and shame the different models on a leaderboard.” An eval that catches on, she says, gives the whole industry a hill to climb together.
“I feel like, you know, the magic happens in these margins.”

North star: AI that helps cure every disease
Paige’s north star is for AI to help cure all diseases and to teach us science that might otherwise stay locked away for thousands of years. She calls AI applied to the sciences the most joyful thing she gets to work on.
In Paige’s view, people change fields when they refuse to stay inside one, and the best tools help people stay whole, curious and a little weird. She acts on that belief every day, from the side projects she ships to the communities she shows up for.
The AI Collective exists for the same reason. We are building social infrastructure for the next billion builders, who arrive with a question, a project, and a reason to care.
A Seat in the Hall
She was told her work was a waste of time, kept building, and is now Head of Developer Relations at Google DeepMind. It began with open source, where the work itself became her interview. She treats developers as everyone, evals as a public good, and curiosity as a job requirement.
Paige Bailey has earned her place in our Hall of Fame.
🎙️ Live today on The AI Collective Podcast

Join AJ Green and Catherine McMillan live at 9 AM PT on The AI Collective Podcast. At 9:15 AM PT, Rachel Soh, Global Lead, Partnerships at Google DeepMind and a 0-to-scale builder who has brought products to market at both DeepMind and Google Research, talks about what it takes to get from research to market. At 9:45 AM PT, Justis Mendez, Co-Founder & Managing Partner at OneSixOne Ventures, a pre-seed fund that hosted more than 75 curated events with Fortune 500 executives, exited founders and investors across 13 cities last year, shares how he finds the founders who will do it next.
Explore Google DeepMind!
The AI Collective is built by volunteers across 180+ chapters in 40 countries.
Thank you to the thousands of volunteers around the world who make this work possible. We truly could not do this without you.
🧑💻 About the Editorial Team

About Josh Evans
Josh is a Managing Editor at The AI Collective Newsletter and leads content for The Byte. Outside of AIC, Josh works in Content Protection at Spotify.
About Our Hall of Fame Inductee

About Paige Bailey
Brings data to opinion fights.
Paige is an AI leader with over 15 years of experience bridging cutting-edge machine learning research and tangible business value. She helped define the developer experience for Google's most advanced generative models, including Gemini, PaLM 2 and Gemma, and is a recognized subject matter expert who partnered directly with Sales teams on Google Cloud contracts during her time at Google.
