Kevin Geonhun Lee
Dartmouth College — B.S. Computer Science • Hanover, NH
I’m Kevin Lee, a CS student who loves the intersection of cybersecurity, networked systems, and ML/AI, with hands-on full-stack experience in research and production. I’m drawn to esports infrastructure—where reliability, latency, and security matter most. Outside of code, I compose, paint, and play video games.
Projects
Selected builds with measurable impact.
Designed a medical imaging pipeline processing MRI/fMRI/DTI inputs via FSL (motion correction, registration, ROI analysis) and delivered results to a mobile app. Built an Android visualization layer using WebView and JavaScript medical imaging viewers for slice navigation and overlay rendering.
Improved concussion screening diagnostic accuracy from 78% → 94% through automated preprocessing and feature extraction.
Built real-time sitting posture simulations using reusable Three.js components modeling vertical and horizontal spine inclination. Classified posture states from pressure sensors (FSRs, textile sensors), accelerometers, and gyroscopes using data from 113 student participants.
Developed a browser-based visualization tool for real-time posture feedback and severity classification.
Building a group trip planning app that turns preferences into a shared itinerary. Includes travel-style profiles, group chat with voting, and automated route optimization for destinations and activities.
Goal: make group trips actually happen by reducing planning effort while keeping everyone’s preferences in the loop.
Experience
Recent roles across systems, security, and product engineering.
- Built production React and React Three Fiber interfaces integrating generative AI analytics for Monumetric advertising data, analyzing placement effectiveness, component performance, and reader engagement.
- Designed backend services and REST APIs in Node.js to process large-scale ad impression and interaction data from external partners collaborating with The Dartmouth.
- Worked cross-functionally with design, business, and external vendors to ship data-driven tooling used by editors and operations staff.
- Designed a multi-layer DDoS mitigation research prototype combining real-time traffic sensors (port mirroring, sFlow) with intelligent filtering across Layer 3/4 and Layer 7 traffic.
- Modeled edge mitigation mechanisms including BGP RTBH and BGP FlowSpec policies to drop attack traffic at routers and distributed PoPs before congestion reaches protected services.
- Simulated real-world attack scenarios in controlled virtual environments (“red button” testing), evaluating ML-based detection accuracy and time-to-mitigation tradeoffs.
- Built interactive dental lab simulations using reusable Three.js components to visualize implant procedures and workflows, reducing development effort for new lab modules.
- Improved an internal clinical support chatbot using retrieval-augmented generation (RAG), few-shot prompting, and structured reasoning pipelines in Python and LangChain.
- Developed full-stack integrations between 3D simulation interfaces and ML-backed services, enabling clinicians to explore procedural visuals alongside AI-generated guidance.
- Supported students developing C programs in Linux-based environments.
- Debugged memory management, file I/O, and socket-based networking issues.
Coursework
Core CS foundations + current focus areas.
Skills
A stack aimed at shipping + performance + reliability.
Contact
Reach out — I reply fast.