Experience

Work

Developing Partner AI to blend compensation strategy, data, and policy into an AI-driven recruiter assistant that delivers context-aware, real-time market-based offer recommendations.
Building LLM guardrails with deterministic software to prevent hallucinations.
Tech led an AI classifier that classifies customer’s jobs into Compa’s job hierarchy with 83% accuracy (baseline accuracy was 70%). This helped decrease customer onboarding time by hours.
Tech led Employee Benchmarks, a product that enables comp analysts to compare their employee compensation to the market.
Tech led an internal tool (frontend + backend) to track data quality issues, akin to Sentry.io. Worked with XFN partners to establish baseline rule set.
Tech led a PII (personally identifiable information) removal system to prevent PII from entering the database.

Worked on Kubernetes backend infrastructure.
Integrated Kubernetes with FluxCD to move to a GitOps workflow.
Made application deployment completely automated with manual promotion from staging to production.

A part of the internal Network UI team.
Delivered an internal UI to enable network engineers to view and test various network topologies quickly.
Worked with various technologies, such as React, Hack, Thrift, GraphQL.

Worked on integrating Google Cloud AI's Platform Optimizer Product with JupyterLab.
Enabled users to easily tune their machine learning model hyperparameters within the JupyterLab environment.
Worked with user experience designers, product managers, and backend engineers to deliver the first full feature user interface for the Vizier Optimizer service.

Led a team of five to organize HackCU 9, a 200 person hackathon. Managed finances, sponsors, judging and logistics.
Led a team of ten to organize HackCU 8, a 150 person hackathon.
Led website redesign and creation.
Helped organize HackCU VI, a 400 person event hackathon.
Created live schedule website to help hackers get information and see event schedule.

Programmed and managed an organization of 8 members which maintained Fairview High School's student-run website, written with Ruby On Rails and has over 400,000 lines of code and nearly 5,000 commits; on average it served over 100,000 weekly page views from students, staff, and parents.
Developed a content creation system (akin to Weebly or Squarespace) that helps our counselors create helpful and engaging content for students and parents; collaborated with counselors.
Created Silent Auction, a responsive web application written for the school's National Honor Society used by parents and students to bid on charity items.