Sometimes I'm curious; sometimes I'm paid. Both result in code.

A mix of ventures, research, and industry experience.

Ventures

  • arcScore.ai

    A NIL valuation and matching platform that makes athlete value easy to understand and easy to act on. Combines performance and marketability signals into one clear score with a realistic earning range.

    arcscore.ai

  • TechnicallyFit

    The operating system for human performance — a global ecosystem where humans and AI collaborate to enhance physical capability through data, coaching, and science.

    technicallyfit.com

  • Chalkless.ai

    Your entire coaching business in one conversation. Describe your coaching practice and Chalkless builds your website, books your clients, processes payments, and manages your business — all from a single AI conversation.

    chalkless.ai

  • Rollout Bayesian Optimization

    Efficient strategies for evaluating and optimizing rollout policies for Bayesian optimization.

    arxiv.org

  • RxStore

    A distributed information system for patient prescription management that is location agnostic, based on Chord, a protocol and algorithm for a peer-to-peer distributed hash table.

    github.com

  • MVBCoin

    A minimally viable blockchain implementation in Python.

    github.com

Industry & Research

  • Cornell University

    Engineering Summer Mathematics Institute Associate with the Office of Inclusive Excellence. Guided 15 students through project-based learning in applied mathematics and designed 10 professional development sessions focused on academic growth, career readiness, and self-advocacy.

  • LinkedIn

    Developed automated time series forecasting methods for hardware capacity planning in LinkedIn's data centers. Identified an intrinsic "level-shift" property in usage data and built regressors for level-shifts, changepoints, and anomalies, achieving a 13.39% improvement over baseline forecasting methods.

  • AMD

    Built neural network-based approaches to reduce inter-node communication overhead in distributed PDE solvers for wave propagation simulations. Led the SHAPE project, creating a C++ system with seamless Python integration for training and deploying ML models in high-performance computing environments.

  • IBM

    Developed local control algorithms for 2D-mesh routing in analog AI hardware, working on next-generation chip architectures to accelerate AI applications.

  • Adobe

    Integrated NLP-based voice and gesture commands for image manipulation, bridging natural language understanding with computer vision in an industry research lab.

  • Stanford University

    Research Intern studying Human-Computer Interaction.

  • Google

    Engineering Practicum Intern. First exposure to industry software engineering standards.