Portfolio
UW–Madison Computer Science and Data Science student focused on AI engineering, software systems, and applied machine learning. I build production-oriented AI systems, full-stack products, automation pipelines, and data infrastructure from prototype to deployment.
3.5+
Years Experience
May 2026 - Present · New York, NY
Built production-grade features for an AI advertising and video-generation platform across Next.js/TypeScript, FastAPI, PostgreSQL, Redis, RBAC, observability, and CI/CD. Architected a four-stage media pipeline that turns three-view product inputs into Gemini-generated keyframes, parallel video segments, and FFmpeg-assembled 720p/1080p output across three animation backends. Delivered a stakeholder dashboard backed by 14 FastAPI endpoints, SSE progress updates, caching, and seven persona-based voice profiles.
May 2025 - Aug 2025 · 3 months · Madison, WI
Supported routine facility inspections and preventive maintenance across campus spaces, using basic tools to address minor equipment and property issues. Documented and escalated maintenance needs to help keep facilities safe, clean, and operational for students, staff, and guests.
Dec 2024 - Feb 2025 · 2 months
Analyzed policy and talent-allocation trends across AI and strategic technology ecosystems for a national-level research project. Collected and reconciled multi-source policy data, compared institutional approaches, drafted research findings, and built a Python linear model to support quantitative forecasting for project stakeholders.
Oct 2024 - Dec 2024 · 2 months · United States
Designed a Python ingestion pipeline with BeautifulSoup, Requests, and Pandas to collect nonprofit tax data from IRS and other public registries. Built HTML parsers and regex filters for filing year, tax category, EIN, and geography, then loaded 50,000+ structured records into PostgreSQL with 98% accuracy to support grant research and partner outreach.
Aug 2024 - Oct 2024 · 2 months · Chicago, IL
Collected, cleaned, and analyzed operational data from databases, APIs, spreadsheets, and web sources using Excel, SQL, and Python. Worked across Microsoft Power Platform—including Power BI, Power Automate, Power Apps, and Microsoft Lists—and Scoro CRM to connect data, support workflow automation, and turn findings into management-ready reporting and process improvements.
Jun 2024 - Sep 2024 · 3 months
Researched the North American Asian food industry by analyzing racial demographics, customer behavior, and consumer preferences with Python and Excel. Produced data visualizations and contributed to a 200+ page market white paper, translating findings into marketing recommendations and multi-platform content designed to improve audience engagement.
LIVE FROM GITHUB
Loading current GitHub data…
Tech:Gemini Live, multimodal voice interaction, dynamic visuals.Impact:Delivered a production-style real-time storytelling demo with turn-based gameplay and live agent responses.
Last updated: 2026-03-16 16:44 UTC
Award-winning project
Recognition:Award-winning group AI dining discovery project.
Tech:Next.js, React, Python, FastAPI, Google Gemini, Google Maps.Impact:Built an AI restaurant discovery product that turns group preferences, location, and natural-language input into explainable recommendations.
Last updated: 2026-04-11
Tech:multi-agent orchestration, desktop workflows, research tooling.Impact:Enabled coordinated agent pipelines for practical research and writing tasks beyond single-prompt demos.
Last updated: 2026-03-14 07:55 UTC
Tech:Python, computer vision, player tracking, pose overlays, shuttle recovery.Impact:Produced stable 2-player badminton visualizations and demo-ready match analysis outputs with more impressive visual storytelling.
Last updated: 2026-03-27
Tech:Python, reinforcement learning, DreamerV3, PPO, experiment diagnostics.Impact:Benchmarked two RL approaches on Super Mario Bros 1-1, documented the full engineering journey, and shipped a PPO baseline that can clear the level with the stronger right-side run shown on the card.
Last updated: 2026-04-27 08:33 UTC
Tech:LLM tracing, latency logging, cost analytics.Impact:Made experiments easier to debug and optimize with visibility into failures, latency, and spend.
Last updated: 2026-03-16 15:33 UTC
Tech:benchmarking, economic tracking, operations dashboarding.Impact:Combined AI coworker workflows and paper-trading operations into one production-oriented interface.
Last updated: 2026-03-19 10:10 UTC
Verified merged PRs across major open-source projects including GitHub, Docker, Microsoft, NVIDIA, Meta tooling, and developer infrastructure. Each link points to a confirmed merged pull request.
RECOGNITION
Selected engineering, data, robotics, and mathematics distinctions. Every card includes a public proof or record link.
Second Place
Built an AI-powered group dining product that matches shared tastes, context, and nearby restaurants.
Venture Track Third Prize
Created an AI rental-evidence workflow that turns leases, checklists, and photos into organized reports.
Third Place
Recognized for a team-built technical solution under a fast-paced university hackathon format.
Second Place
Applied Python-based analysis and data storytelling in a competitive datathon setting.
Robot Skills 2nd · Qualification 3rd
Robotics recognition combining C++ programming, build iteration, and competition strategy.
Verified Winner Badge
Official Devpost achievement awarded after winning a hackathon prize.
Google certificate focused on applying AI for research workflows, insight generation, and practical knowledge work acceleration.
Show credentialDemonstrates data analysis fundamentals with analytics workflows and Python-based problem solving.
Show credentialFoundational AI certificate covering practical use of modern AI tools and core concepts for applied engineering and productivity.
Show credentialValidates practical experience building applications with Anthropic APIs and production-oriented AI integration patterns.
Show credentialFocuses on AI frameworks, foundational concepts, and structured understanding of how modern AI systems are built and used.
Show credentialHighlights understanding of model strengths, constraints, and practical AI evaluation for real-world engineering decisions.
Show credentialAnthropic certification covering MCP fundamentals and how model context is exposed through practical tool protocols.
Show credentialVerified software engineering assessment covering core programming, problem solving, and engineering fundamentals.
Show credentialCredential for building AI-assisted no-code workflows and structured Airtable app experiences.
Show credentialGoogle professional credential demonstrating applied AI fluency across practical tools and workflows.
Show credentialIntroductory Anthropic certification covering Claude workflows, prompting patterns, and agent-oriented practical usage.
Show credentialShows familiarity with deploying and using Claude through Google Vertex AI, including cloud AI integration workflows.
Show credentialOpen to software engineering internships and project collaboration. End a quick message and I'll follow up.
Book a 30-minute meeting directly on my calendar.
