~/jovy
Jovy Zhou
Software Engineer
Education
Northwestern University
BA/MS Computer Science
GPA: 3.92 / 4.00 · Dean's List
About
Full-stack engineer with a weakness for hard problems — distributed systems, LLM pipelines, and the occasional V6 crimp slab. Interned at GitHub on infrastructure that serves 50M+ daily visitors. When I'm not in the terminal you'll find me on the climbing wall, fine-tuning a QK Alice Duo (Osume Mochi switches), or in ranked queue — whichever is currently winning the battle for my attention.
Experience
Software Engineer Intern
GitHub · San Francisco
- ▸Replaced an 8-year-old legacy sticky-header library across 20+ high-traffic Rails surfaces (50M+ req/day) with native CSS position:sticky + lightweight JS utilities
- ▸Drove staged feature-flag rollout across 10+ partner teams with Sentry/Datadog instrumentation — achieved zero regressions and 60% reduction in bug reports
- ▸Built a Copilot Agent-driven Jest → Vitest migration pipeline (API detection, prompt transform, diff validation) — cut manual effort 80% (10m → 2m/file) with 72% clean pass rate
- ▸Produced onboarding docs enabling 500+ engineers to self-serve; pilot migrated 10 packages / 50+ test files with zero production regressions
Teaching Assistant
Northwestern University
- ▸CS 211 — Supported 200+ students in C/C++ (OOP, memory management, debugging); raised assignment completion rates by 20%
- ▸CS 310 — Guided 400+ students deploying AWS serverless services (Lambda, API Gateway, RDS/S3)
Projects
Ripplet
MU Collective
AI / EdTechSep 2024–Present
- ▸AI exam authoring platform (Next.js, PostgreSQL) cutting draft time 45m → 15m via LLM prompt templating + refinement tools
- ▸Optimized LLM usage via prompt caching and batched generation, stabilizing cost as usage expanded
- ▸Conducted 30+ user interviews; prioritized weekly releases achieving 80% user retention (rubric editor, tagging)
Next.jsPostgreSQLOpenAI APITypeScriptPrompt Engineering
Full-StackSep 2024–Present
- ▸Office hour queue platform for 500+ students (React, Firebase) with reliable real-time updates
- ▸Intelligent question grouping reducing duplicate handling and shortening wait times by 25%
- ▸Multi-tier priority queue ensuring students still waiting for initial help were served first
ReactFirebaseRealtime DBTypeScript
CTEC Snap
🏆 2nd place — Wildcat FreshHacks
AI / ToolingOct 2024
- ▸Shipped MVP in an 8-hour sprint with a 5-person team
- ▸React frontend + Flask API + LLM pipeline (batch chunking → summarization → topic tagging)
- ▸Surfaces thematic summaries from course evaluation text to help students make informed course decisions
ReactPythonFlaskLLM PipelinesPrompt Engineering
Skills
Languages
TypeScriptJavaScriptPythonJavaC/C++RubySQL
Frontend
ReactNext.jsTailwindGraphQLREST
Backend
Node.jsFlaskRailsSpring BootPostgreSQLMongoDBRedis
Infra & Tools
DockerGitHub ActionsAWSFirebaseSentryDatadog
Focus Areas
Distributed SystemsSystem DesignAI / LLMFull-StackScalability