Storrs, CT · B.S. Computer Science & Engineering + B.S. Mathematics, UConn

Luis Mendez

Building at the intersection of AI/ML, quantum computing, and applied security — currently researching computer vision at UConn and shipping hackathon projects that keep placing.

Email me GitHub LinkedIn

01About

I'm a dual-degree Computer Science & Engineering / Mathematics student at the University of Connecticut, concentrating in AI, and expecting to graduate in April 2027. My work spans two threads: research — pose-based person re-identification, transportation infrastructure detection with Vision Transformers, and stochastic process simulation — and fast-moving hackathon builds in quantum computing and Bitcoin/Lightning security, several of which have placed first. I'm active in Husky Developers, SHPE, ColorStack, AI Club, the Quantum Computing Club, and Codepath.

02Experience

  1. Undergraduate Researcher — Person Re-Identification

    May 2025 – Present

    Dr. Yuan Hong Research Lab, UConn · Advisor: Dr. Hanbin Hong

    • Developing a pose-based Person Re-Identification (ReID) model for robust identity matching.
    • Exploring keypoint refinement, Brownian motion augmentation, and Vision Transformers to improve embedding consistency.
    • Implementing and benchmarking model variants for accuracy and robustness under occlusion and pose variation.

  2. Transportation Systems Research (CTDOT Collaboration)

    2025 – Present

    Dr. Yuan Hong Research Lab, UConn · Advisors: Dr. Hanbin Hong, Dr. Kai Wang

    • Developed a Vision Transformer model for detecting centerline rumble strips from roadway imagery.
    • Built a scalable computer vision pipeline for automated infrastructure detection on large-scale datasets.
    • Designed data pipelines for frame extraction, sampling, and annotation across multi-terabyte roadway archives.

  3. Research Assistant — Department of Mathematics

    Sep 2025 – Dec 2025

    University of Connecticut · with Graduate Assistant Albert Michael

    • Simulated Brownian motion and stochastic differential systems for applied probability research.
    • Developed Python-based simulation and visualization tools for random processes and diffusion phenomena.
    • Performed numerical verification of theoretical models related to stochastic convergence behavior.

  4. Software Engineer Intern — Valour Software (Open Source)

    Sep 2024 – Jan 2025
    • Contributed 10,000+ lines of code to a community-driven platform using Git and agile workflows.
    • Developed a user engagement feature enabling reactions via RESTful APIs and a modular frontend.
    • Extended the database schema using ModuleBuilder, supporting 5,000+ active users.

03Projects

Aegis Wallet

MIT Bitcoin 2026 · Top 10

Bitcoin · Lightning (LND/litd) · MCP · Node.js · Next.js

  • Secure MCP server exposing 7 wallet tools for agent-driven transactions with macaroon-based authentication.
  • AI-agent wallet interface enabling autonomous transaction execution without exposing private keys.
  • Lightning Network integrated in Docker with account-based isolation and scoped permissions.

Quantum Risk Engineering

MIT iQuHACK 2026 · 1st Place

Qiskit · NumPy · SciPy · Matplotlib · Classiq

  • Benchmarked Monte Carlo vs. Iterative Quantum Amplitude Estimation for VaR and CVaR.
  • Demonstrated quadratic error-scaling improvement of quantum methods across precision levels.
  • Proposed Soft-CVaR, cutting quantum circuit depth ~5x and improving NISQ feasibility.

Full Quantum Hash Function

Yale · 1st Place

Qiskit · NumPy · Matplotlib

  • Designed a quantum-secure hash function using quantum walks and parametric circuits.
  • Encoded classical inputs into entangled quantum states for fixed-length outputs.
  • Validated entropy, determinism, and collision resistance on IBM's 20-qubit simulator.

Phoenix

HackHarvard 2025

Google Cloud Vision (Gemini) · PyTorch · React · Flask

  • AI platform leveraging Google Cloud Vision to enhance document intelligence.
  • Visual similarity system using Vision Transformers and cosine similarity.
  • Full-stack application for real-time feature extraction and analytics visualization.

Neural Data Visualization Tool

HackDartmouth

C++ · ImGui · CMake · Python (SciPy, NumPy)

  • Real-time C++ application to visualize spike-based neural signals.
  • Applied bandpass filtering and normalization to raw waveform data.
  • Implemented multithreaded rendering for low-latency performance.

04Skills

Languages

PythonC++Java CC#JavaScriptTypeScript

Backend

Node.jsExpressFlask REST APIsWebSocketsDocker MCPGitLinux

Frontend

ReactNext.jsTailwind CSS AngularVue.js

Machine Learning & CV

PyTorchVision TransformersOpenCV Hugging FaceScikit-learnNumPy PandasSciPyLangChainLangSmith

Blockchain & Security

BitcoinLightning Network Post-Quantum CryptographyZero-Knowledge Proofs

Certifications

Microsoft Foundational C# Frontend Masters — Functional JS & Angular edX — Python for Data Engineering

05Education

University of Connecticut

Apr 2027

Dual Degree: B.S. Computer Science & Engineering — B.S. Mathematics · AI Concentration · Mansfield, CT

Husky Developers · SHPE · ColorStack · AI Club · Quantum Computing Club · Codepath