At Facebook, I worked on their Network Infra team, improving the performance and reliability of a service that generates network device configs.
The service handled ~3k requests/hour. My work saw a ~10% performance increase, and prevented several SEVs.
At JPMorgan, I worked on data modeling, and data engineering at scale.
This looked like 12,000+ data points with 40+ properties on an hourly basis.
Dealt with constantly changing business requirements from investment bankers, varying data processing issues (asynchronous fetching on dynamic time cycles), etc.
At Sihua Technologies, I improved the deployment time for their CI/CD pipeline by 70%, but the tool was completely unmanagable, the code was absolute crap, and not used in production.
I also run an app development club at Rice called RiceApps. We're about 50 members strong now, and we're building some pretty ambitious projects this year.
At Rice, I've served as a TA for a few CS courses (head TA for COMP 322 Parallel Programming, and COMP 330 Data Science) and had an unfortunately short and hardly significant undergrad research stint in deep learning.