~ Take me home

June 2024 to now | Traveling

I'm taking some time off to travel, work on personal projects, and chart out what's next. Stay tuned!

January 2023 to June 2024 | Snowflake

After Myst was acquired, I spent a year and a half working to improve Snowflake's machine learning capabilities.

During my first few weeks, I collaborated with the team to share our expertise in developing production machine learning systems. This involved sharing knowledge about our APIs, time series database, model training and inference platform, and hyperparameter tuning services.

I then worked on a nimble team to design and implement Snowflake's first semantic search functionalities. We shipped early access support for built-in vector distance functions, created services for approximate vector search and text embedding generation, and contributed to Frosty, an LLM-powered chatbot built for Snowflake. These efforts helped pave the way for native vector data types, embedding generation, and search.

Motivated by a desire to learn more about performant and secure systems programming, I transitioned to the Snowpark Execution Environment team. This team was responsible for the platform that powered many machine learning features like Snowflake ML and ML Functions. I focused on optimizing Python environment creation for interactive workloads like Streamlits and Notebooks, and improving support for memory and compute-intensive workloads like distributed hyperparameter optimization.

September 2018 to January 2023 | Myst AI (acquired by Snowflake)

In September 2018, I joined Myst as the founding engineer. For four and a half years, I sharpened my software engineering and machine learning skills, and helped build and scale two SaaS products.

I worked on all parts of our technical stack. Some of the projects the team shipped that I'm most proud of include our networks-of-time-series data model, relational and time series data persistence layers, model graph building experiences, REST APIs, clients, modeling libraries, model training and inference platform, model definition libraries, and backtesting and hyperparameter tuning services.

I spent most of my time programming, leveraging a wide array of technologies, including: machine learning techniques and libraries like Bayesian optimization, autoregressive modeling, Numpy, Pandas, Ray, XGBoost, and Scikit-Learn; infra like Kubernetes, MySQL, PostgreSQL, BigTable, Spanner, and Pub/Sub; and libraries for building web services like Flask, FastAPI, and React.js.

In addition to my technical work, I contributed to our culture, engineering roadmaps and processes, hiring, mentorship, and team-building.

I also did some apprenticeships with coworkers to round out my data science and business development skills and learn how other parts of the business functioned. There was no shortage of work to be done at Myst so I tried to contribute where I could.

August 2015 to September 2018 | Undergrad & Internships

I took a non-traditional path through my undergraduate studies that gave me a wide range of academic experiences. I started at Carnegie Mellon University studying Biomedical Engineering, attended a community college in the Bay Area, and then transferred to Pomona College where I studied Computer Science and Cognitive Science & Linguistics.

I played Division III soccer for Carnegie Mellon and Pomona, which took most of my time during the fall semesters. During community college, I was a math tutor, member of student government, and hosted a podcast for the school newspaper. At Pomona, I was an associate at Pomona Ventures, a student-run entrepreneurship club, helped organize the 5C Hackathon, and did language topic modeling research with a professor.

In the summer after my freshman year, I interned at Google Nest. I worked with one of the future Myst co-founders to build an ETL framework with Airflow, Dataflow, and Kubernetes to evaluate the performance of the Nest Home/Away Assist algorithm across devices in the wild. Eventually, the framework was adopted by multiple teams to scale their data pipelines.

In the summer after my sophomore year, I backpacked with friends through South East Asia.

In the summer after my junior year, I returned to Google Nest, and worked on a simulator to test Smart Home algorithms.