Ideas from the frontier of Applied AI
Thoughts on screen-aware AI, multimodal models, privacy-first design, and building technology that actually helps people think.
First posts coming soon
We're a team of researchers and builders. We have a lot to say — about AI architecture, privacy, the state of multimodal models, and what real screen-aware AI looks like under the hood. Watch this space.
Topics we'll cover
Applied AI research
How multimodal models work, where they fail, and what comes next. Written by people who read the papers.
Screen-aware interfaces
The design and technical challenges of building AI that understands visual context in real time.
Privacy & AI
How to build powerful AI products without sacrificing the trust and autonomy of your users.
Prompt & context engineering
Practical techniques for getting the best out of large language and vision models in production.
Building in public
What we learned shipping HeyScreen — technical decisions, wrong turns, and what we'd do differently.
AI for students & researchers
How doctoral researchers and students can use AI tools to accelerate their work without compromising integrity.