Collaborated cross-functionally in an ambiguous environment to build 15+ greenfield prototypes/MVPs. Utilized low-code tools like Retool, Airtable and built full stack applications powered by public APIs or custom ETLs/APIs.
Languages/Frameworks/Tooling: Python, Django, Flask, Pandas, Pytest, Scrapy, SQL, SQLAlchemy, Alembic, Next.js, React, Mapbox, Tailwind CSS, Postgres, Airflow, Docker, Kubernetes, GitHub Actions, AWS, Vercel, Mapbox, LangChain, OpenAI, Zyte, Jupyter Notebook, Airtable, Retool, Notion, JIRA, Figma, Sentry, Mixpanel, and more.
•Led intern hiring and onboarding; managed 2 interns, both secured full-time engineering roles post-graduation.
•Led search feature rebuild for an internal product, migrating from Amazon Kendra to Elasticsearch, improving data quality and refactoring ETL and APIs, resulting in a 100% increase in user trust in search results.
•Tech lead, architect and backend developer on WorkInEd, a jobs search site collating DOE teaching roles:
> Designed and implemented a scalable backend for scraping 13k+ public school districts, supporting shared scrapers and integrating Airtable to help the product manager track and manage school district data.
> Built ETL pipeline to dedupe, clean, and enrich scraped data ensuring accurate and up-to-date postings.
•Core backend engineer on Immigration Commons, a closed platform providing reliable immigration data:
> Implemented API endpoints for summary visualizations of 10+ years of EOIR data, enabling employees from 13 state offices and 30 national organizations to leverage time series insights and inform advocacy efforts.
> Refactored core (EOIR) data ETL and integrated OpenSearch index and Vercel cache resets to fully automate ETL process, fully removing manual engineering effort and improving reliability.
> Scraped state immigration data and prototyped RAG chatbot extension, validating feasibility of new feature.