Working on data quality, data classification, and agent quality

Worked to make Tecton the de-facto solution to build production ML feature pipelines.

Led efforts to build out Rift (https://resources.tecton.ai/rift), the next-gen Python based feature compute engine.

Managed the Feast (http://feast.dev) team to improve the best-in-class open-source feature store, growing project adoption by ~5x.

Drove improvements to post sales processes while directly owning and expanding contracts for some of Tecton’s largest customers.

Worked across ML research teams and product teams to make Google Docs/Sheets/Slides intuitive and intelligent consumer and enterprise product offerings across web and mobile platforms.

Included building models (incl LLMs), working with research orgs, building a privacy-aware ML infra layer, and integrating with products to launch features to production.

• Built features to improve student engagement with educational content in Tanzania.

• Designed and implemented a custom search engine.

• Co-designed a database system to work with content consumers and editors across languages.

• Worked with team of designers to build a performant Android survey/location tracking application to enable information crowdsourcing and urban planning research in developing countries. Deployed across projects in Mexico, Bangladesh, Indonesia, etc. (http://www.flocktracker.org/)

• Developed responsive web pages to distribute Dhaka's first public bus map to streamline transit and reduce congestion/pollution.

• Prototyped Google Maps-based dashboard, allowing companies in Dhaka to manage their bus fleets.

• Mentored an engineer to learn design principles and eventually take over the project to continue iterations and deployment.