Where I keep my talks, writing and projects. Machine learning, data engineering, AI, and the occasional life lesson.
Looking for guidance or help with a build? I offer consulting for founders, websites and applications, AI models, and homelabs.
Now
Monster Mash 2 · Live demo
Monster Rally
The monsters moved to San Francisco and started Monster Rally. Drop by their Halloween office party, powered by Temporal and Pydantic AI. Open “See Temporal live” to follow the model calls.
Explore the live demo-
Temporal: Durable workflow platform for building reliable AI agents. My demos, resources and latest work.
-
Vibe Check: My ongoing YouTube series on AI tools, agentic coding, and building AI in production
-
Fight Health Insurance: Co-founded startup using AI to help people appeal denied healthcare claims
-
GitHub: Code and projects
-
Tech Blog: Deep dives into data engineering, machine learning, AI and Google Cloud Platform
-
LLM Fine-Tuning: Eight-part series taking a model from “should I even fine-tune this” to running in production, each post with video and audio
Talks & Media
-
Technical Talks: Conference presentations and talks on AI and ML topics
-
Speaker Decks: Slide decks from the various tech talks above
-
GCP Podcast: Co-hosted the Google Cloud Platform Podcast, nearly 60 episodes covering AI, ML and developer tools
Resources
A few of the trackers and news sources in my rotation for keeping tabs on tech and AI.
-
Tech Brew Ride Home: Brian McCullough’s daily 15-minute tech news rundown (formerly Techmeme Ride Home)
-
PlushCap: Matt Makai’s developer trend tracker. Content analytics and rankings across 500+ devtools companies, plus trend signals on MCP and AI agents
-
Artificial Analysis: independent LLM benchmarks for intelligence, speed and pricing across models and providers
-
LMArena: Community-voted head-to-head leaderboard for AI models
-
Hugging Face Trending: What’s rising in open models right now
-
Epoch AI: Independent research and data on AI trends: compute, model capabilities, chips and data centers
Archive
-
The OCEAN mailing list data set: first author on this 2022 paper, network analysis spanning mailing lists and code repositories, from a Google and University of Vermont research collaboration
-
Vermont OCEAN: the project it came out of, Open Source Complex Ecosystems and Networks
-
Deeplearning4j: the open source neural net platform I helped build at Skymind, 2015-2016
-
ND4J: the numerical computing library underneath it
-
DataVec, dl4j-benchmark, dl4j-model-z, dl4j-0.4-examples and ImageNet-Example: the rest of the set
-
Why these are archives: the originals were deleted in the 2019 Eclipse monorepo consolidation, which rewrote history and dropped contributor attribution, including roughly 1,900 of my commits. I recovered these from surviving public mirrors.
-
Techtonica: Former Board Director (2019-2021) supporting programs to increase female and non-binary representation in tech
Enjoying the content? Contribute to the chaos ☕