Introduction to Programming
Programming fundamentals in Python: control flow, functions, memory diagrams, and unit testing.
About Me
I'm a junior studying Computer Science at UNC Chapel Hill. I like building things and figuring out how to make them actually useful.
I'm currently doing applied AI research at UNC, working on MAIA, an AI programming tutor, and I've previously worked in enterprise software and helped start a student-run credit union.
When I'm not coding, you'll probably find me watching Carolina basketball, working on some random project, or getting coffee on Franklin Street.
Classes I've Taken
Coursework I've completed at UNC, plus what I'm taking now.
Data Structures & Analysis
Linked lists, trees, hash tables, and graphs, with asymptotic analysis for comparing implementations.
Systems Fundamentals
C programming, pointers, and manual memory management, along with how source code gets compiled, linked, and executed.
Discrete Structures
Propositional logic, proof techniques including induction, combinatorics, and introductory graph theory.
Foundations of Programming
Object-oriented design in Java: inheritance, interfaces, abstraction, and common design patterns.
Computer Organization
Digital logic, assembly language, and processor design, including pipelining and the memory hierarchy.
Software Engineering Lab
A semester-long team project for an external client, covering requirements gathering, sprint planning, and code review.
Multivariable Calculus
Calculus in several variables: partial derivatives, gradients, vector fields, and multiple integrals.
Modern Web Programming
Front-end and full-stack web development.
Skills
Languages, frameworks, and tools I've worked with.
Languages
- Python
- Java
- TypeScript
- JavaScript
- C++
- SQL
- HTML
- CSS
Frameworks & Libraries
- React
- Next.js
- Node.js
- FastAPI
- PyTorch
- scikit-learn
- Matplotlib
- AJAX
AI & Machine Learning
- RAG
- Vector Embeddings
- HNSW Indexing
- MCP
- Prompt Engineering
- LLM Evaluation
Developer Tools
- AWS
- Docker
- CI/CD
- REST APIs
- Microservices
- Git
- Linux
- Supabase
- Tailscale
- Agile
Experience
Research, internship, and student organization work.
Applied AI Researcher
UNC Department of Computer Science
Jan 2026 – Present
- Designed MAIA, an LLM-powered programming assistant that holds 92.5% adherence to Socratic tutoring principles through a custom ruleset that prevents direct code generation.
- Constructed a 388-turn evaluation dataset scoring 0.73 Cohen's Kappa reliability, built on an LLM-assisted classification pipeline over live Java debugging sessions.
- Scaled the tutor to 300+ undergraduates by deploying its microservices architecture across a university-level course.
Software Implementation Consulting Intern
Intapp · Enterprise CRM
Jun 2026 – Aug 2026
- Configured 50+ custom CRM objects from technical discovery findings and demoed the live platform to executive stakeholders, landing 3 successful implementations.
- Developed an AI workflow using Claude browser skills to parse the DOM, cutting data migration time by 75%.
- Architected a key-person risk identification playbook for the Celeste AI platform, adopted by 3 investment banks.
Co-Founder, Outreach
Carolina Students' Financial
Apr 2025 – May 2026
- Raised $1.2M+ in capital from 10+ organizations for a peer-led credit union built around financial accessibility for students.
- Negotiated a charter partnership with Wasatch Peaks Federal CU to enable federally insured accounts for members.
Projects
Personal projects built outside of coursework.
August 2026
Semantic Caching Layer for LLM APIs
A caching layer that serves semantically similar LLM requests from earlier responses instead of re-running inference. I implemented the HNSW vector index from scratch; it reaches 99.5% recall@1 at 1,000 vectors and runs 1.5× faster than brute force at 10,000. The similarity threshold is calibrated against a 194-pair labeled evaluation set using an asymmetric cost model, which holds the wrong-match rate at 7%. Deployed on AWS with CloudFormation, least-privilege IAM, and budget-triggered shutdowns.
View on GitHub →April 2026
LocalMind
A self-hosted LLM stack running on a home RTX 3060 Ti, built on Ollama, Open WebUI, and Docker, and reachable from authenticated devices over a Tailscale VPN. A two-way sync keeps OneDrive files indexed in a Chroma vector store for retrieval, using event-driven watchers with 2 to 5 second latency and a five-minute reconciliation loop with MD5 state tracking to recover missed syncs.
View on GitHub →October 2025
GravityFit Exo
A Next.js and FastAPI web app that recommends workout intensity based on the surface gravity of a given exoplanet. The recommendations come from linear regression models trained in scikit-learn on 33,000+ NASA exoplanet records, validated with 10+ visualizations of the gravity-to-intensity mapping. Placed first out of 20+ teams at the Carolina Data Challenge.
Visit the site →