I joined the Delta Lives Table team where I worked on data sample preview, building a new end to end feature for sampling large datasets processed and transformed in data pipelines. I leveraged internal Spark APIs to overcome data sampling for streaming queries (incrementalization) where data isn't reprocessed. I orchestrated a new framework that enabled me to expose data transformed on-the-fly to external modules, extending this data sample project to support dataset not persisted as well as further use cases to improve observability and user debugging experience. I productionized the project by implementing event truncation to improve service stability and was able to ship it to production.

Research project, RALF, led by Sarah Wooders and advised by Professor Joseph Gonzalez on improving real time performance of feature stores. I contributed to the open-sourced RALF client, enabling writing feature transformations and querying target tables and load-tested various database connectors to optimize query latency for existing workloads. To support latency and accuracy benchmarks, I designed and built a benchmarking framework that ingests raw data from Kafka streams with sharding policies, leveraging Ray’s distributed data processing with autoscaling kubernetes cluster on AWS EC2 instances.

https://github.com/feature-store/ralf

I joined the Application Discovery Team under the Migration services for AWS, working on a customer facing feature for the ingestion workflow. I designed and built a deployment infrastructure for database schema upgrades for resources in 3,000+ existing customers’ accounts for additional data delivery streams. This project integrates automatic deployment into existing services to support immediate updates of new message types, reducing dependencies between teams to push changes into production while ensuring backward compatibility for existing users. During my time at AWS, I've learned to design more scalable systems that reduce stress on both development time and resource consumption.

I joined Jubily as a fullstack web development intern, primarily working on the search and profile pages for the healthcare providers. I worked on optimizing search queries and management of healthcare providers' registration and profile information.