The focus of my role at Lyft was thinking about the in-vehicle compute for the self driving car project. I'd like to talk about what I built, but my exit paperwork had language just vague enough to discourage mentioning too many specifics. I think it's safe to say that I co-lead the design, bring up, and maintenance of the next generation 'big' computer that that sits in the trunk. Think: many GPUs, server grade Xeons, and enough thermal paste to paint a wall. Automotive requirements, built like a research endeavour. Tempered excitement, lots of avocados. I also spent a lot of time evangelizing neural network acceleration hardware and software techniques, which is what I left to focus on.