daniel's blog
home • blog • projects • contact
Avatar

👋 Hi, I'm Daniel! I'm a machine learning engineer with an emphasis on the intersection of research and product.

I'm currently at xAI, working to advance the best AI systems in the world.

I've spent time in both the Computer Vision and Natural Language Processing domains, most recently as the Head of AI Engineering at Harvey AI. Before that, I did Computer Vision Research at Tesla on the Autopilot team, helping build and ship Full Self Driving. You can see more of my work on LinkedIn.

This website exists for three main reasons:

1️⃣ Firstly, as a representation of my career and resume. If you came hoping to understand who I am and what I've worked on, welcome! You can find my resume here.

2️⃣ Secondly, as a public repository of my writing. I am a strong believer in writing to solidify my thinking and building clarity by formulate ideas onto paper. Doing this in public is an extra forcing function to be rigorous in my thinking, so I try to do it when time permits.

I write primarily about interesting product development and engineering challenges. The psychology of building a great product, the organizational challenges of scaling startups, the latest research in ML, and more. I tend to write in a build-in-public style, so some of my posts are more polished than others. I try to share my learnings as I go, writing what I wish I knew years ago.

3️⃣ Thirdly, as a sandbox to tinker with things I'm interested in. As the cost of software drops to zero, the pipeline of idea → software prototype → ship to the internet is increasingly low friction. If you hold the rough mental model of a web browser as "a way to let other people easily run the code I write", this makes tinkering lots of fun! More-or-less anything that doesn't require authentication or heavy compute processing can be written as a client-only application, aka a static site.

I'm not a web developer by trade, but I enjoy experimenting. (I've been quite interested in local-first LLM inference lately for example, and have been playing around with WebGPU to see how well you can run small models in the browser). These sorts of small utilities aren't usually deeply novel or groundbreaking, but are written with some interesting artistic perspective in mind, or specific use case that I personally care about.

Welcome to my site! Writings, projects, and contact info below.

Latest Posts

See all posts
  • My Learnings and Beliefs
    A collection of micro thoughts
  • Where to apply generative AI (or, why Harvey succeeded*)
    Ways to correctly and incorrectly apply generative AI
  • CV vs NLP
    How to move from Computer Vision to Natural Language Processing

Recent Projects

See all projects
  • 3D Embedding Visualizations
    Interactive 3D visualization of text embeddings using AI and dimensionality reduction
  • In browser LLM inference
    Running a small LLM in the browser
  • Experimenting with Cloudflare
    Proof of concept of Cloudflare functionality

Contact Info

If you want to get in touch with me about something or just to say hi, reach out on social media or send me an email. I'm always open to meeting new people and interesting conversations.

  • X
  • GitHub
  • LinkedIn
  • Resume (PDF)
  • daniel@hunter.sh
© 2025
about this website
Built with Astro and Cloudflare Workers
Click for more info