I managed the Compute team, which was responsible for the compute fleet for Lyft's workloads. My main project was Lyft's migration to Kubernetes. I was rated one of Lyft's best managers in 2019. :)

Infrastructure at OpenAI covers a wide spectrum, from structuring our cloud and on-prem resources, scaling the experiments and clusters, all the way up to the experiment launching frameworks and monitoring tools .- As the first engineer, I designed and scaled our cloud/colo systems. This included working with hardware/hosting vendors and contractors.- I optimized our Kubernetes cluster with a custom autoscaler (https://github.com/openai/kubernetes-ec2-autoscaler).- We wrote about early versions of our infrastructure in a blog post (https://blog.openai.com/infrastructure-for-deep-learning/), and I gave a keynote at KubeCon about how we leverage Kubernetes.- I was responsible for our migration from AWS to Azure.- I have fun stories about scaling up Kubernetes and debugging the network stack. :)- I also spent time on managing, recruiting, designing our engineering interviews, and leading our diversity and inclusion efforts.

- Worked on the core API platform and infrastructure- As a member of the founding team, I implemented engineering processes to onboard/mentor junior engineers

- Led the development of the Incubator project (incubator.duolingo.com), a platform for crowdsourcing language courses. The incubator enables the development of courses by providing a web tool for contributors and staff, maintains all the data used to generate lessons and grade answers, as well as provides data access APIs for other teams in the company.- Was responsible for building a localization tool to support crowdsourced translations, Android app, lesson generation engine, A/B testing framework, web frontend, original speech recognition engine, etc.

15-451 Algorithms