I often get asked “What do you actually do?”. So i'm going to start sharing a bit more detail on how my Startups Applied AI team at @OpenAI partners with frontier startups & shapes OpenAI TLDR - my team has a dual role: ensuring startups get the best out of our models/products, and being the bridge between frontier startups and OpenAI’s research & product teams. I’ve seen sprints result in startups saving >25% of their LLM costs ($ millions), improve latency or eval results materially, and even reducing the startup’s customer cancellations by 40% - not bad for direct impact. And then there’s the feedback that shapes OpenAI’s roadmap. This is a first (high level) post of a few i'll make over the next few weeks, getting much more in depth on various topics. LMK if there are any particular topics you want me to go into depth on!