Max Buckley
Currently exploring founding my own startup — watch this space
About Me
I’m Max Buckley, a software engineer and AI researcher with 12+ years of experience building software and ML systems at scale.
Right now I’m exploring founding my own startup — watch this space. For ~5 months I headed up Knowledge Research at Exa, where I also founded and led the Zurich office. Before that I worked for 12.5 years at Google across various European offices, where I worked on all sorts of things, but ended up in information retrieval, knowledge graphs, and large language models. My work lives in that space where structured knowledge meets the messiness of real-world text—figuring out how to make AI systems that actually know things, not just predict tokens.
I’m based in Zurich, where I live with my wife Alena Nikolskaia. I’m a part of the growing European AI research scene. I write about AI on LinkedIn (where some people apparently find my research synthesis useful), and co-host a podcast, Signal and Stories (more on that below).
When I’m not reading papers or arguing with build systems, you’ll find me in the gym working my way back to doing heavy deadlifts. I’m always interested in conversations about retrieval systems, knowledge management, the economics of AI tooling, and what it actually means to build useful things in this strange moment we find ourselves in.
Signal and Stories
Signal and Stories is the podcast I co-host with Boris Meinardus. Since February 2026 we’ve been sitting down for long-form conversations with the engineers, researchers, and founders shaping AI, cutting through the hype to the technical detail and the human story behind it. Guests so far include Nicolay Gerold on building the Amp coding agent, Emilio Andere on chipping away at NVIDIA’s CUDA moat, and Ben Lau on leaving a McKinsey partnership for forward deployed engineering at Cognition.
Listen on YouTube, Spotify, Apple Podcasts, or Amazon Music. If you’re working in or around AI and have an interesting career story, we’d love to host you!
Interactive
I occasionally build interactive explainers for ideas that are easier to see than to read about. Find them all in Labs:
- How Decision Trees Carve Up Space — a build-it-in-your-browser tour of how decision trees classify points, and why averaging them into a random forest produces a smooth boundary.
- Finding the 1%: Trees & Forests on Imbalanced Data — why a 99:1 class split makes models ignore the rare class, and how weighting, RUS/ROS bagging, and SMOTE fix it.
- Trees That Fix Their Own Mistakes — gradient boosting, stochastic row/column sampling, and why tree ensembles interpolate beautifully but can’t extrapolate.
Connect
Want to get in touch? Find me on: