A ton of stuff, arguably too many things. Right at the moment, my main focus is redoing fundamental 2D vector graphics geometry operations (stroking, path intersection) to make them more robust and higher quality. Some of this will also run in the GPU as compute shaders, which adds some interesting constraints and ups the difficulty level.

I'm also building a new, high performance, high quality 2D rendering engine that runs primarily on the GPU.

And I'm also exploring new approaches to building GUI, primarily in Rust. A new focus is to integrate that more tightly with the renderer mentioned above. A benefit I expect is that I'll be able to do the CPU side of the rendering work in multiple threads, where existing architectures make it very difficult to break out of a single thread.

How did I end up working on this? It's a natural evolution of what I've been working on for a long time, in Ghostscript, on Google Fonts, working on the text stack on Android, and now fortunate enough to have a research position back on Google Fonts where I am able to focus on this and have great support from the rest of the team.