Founders Toolkit

• RL Fine-Tuning of Language Models

  Research into reinforcement learning techniques for shaping language model behavior beyond supervised training. Explores reward modeling, policy optimization, and preference alignment as tools for building more reliable AI systems.

  Reinforcement Learning Language Models Fine-Tuning Jupyter
  Active GitHub ›

• Elo Refined Curriculum Training + RLOO

  Stanford CS research with Donnie Raymond extending RLOO with curriculum learning. Problems and the agent are scored with an Elo rating system, and a Gaussian scheduler samples problems matched to the agent's current ability so difficulty ramps as the model improves. Built on the same RLOO training setup with extensive Elo metadata logging, evaluated on pass@k against SFT and vanilla RLOO baselines.

  Reinforcement Learning Curriculum Learning RLOO Elo Rating LLM Reasoning
  Research GitHub ›

• Swish Backend

  Production Python backend service. Server-side foundation for a consumer application, covering API design, data modeling, and service architecture.

  Python Backend API Design
  Active GitHub ›

• Decision Making Under Uncertainty

  Final project for Stanford CS238. Applied probabilistic decision-making frameworks including MDPs, POMDPs, and policy search to a structured real-world problem, benchmarking solution quality across methods.

  MDPs POMDPs Policy Search Python
  Research GitHub ›

• Strabismus Baseline Classifier

  An academic research project: a baseline machine learning classifier exploring detection of strabismus from clinical data, establishing a reproducible performance benchmark in ophthalmology. This is research code only; it is not a medical device, has not been cleared or approved by the FDA or any regulator, and is not intended for clinical use, diagnosis, or treatment.

  Medical AI Classification Computer Vision Jupyter
  Published GitHub ›

• Logistic Regression for World Football XI

  End-to-end ML pipeline selecting an optimal World XI from player performance data across global leagues. Covers data preprocessing, model training, predictive analytics, and player selection based on logistic regression outputs. A demonstration of principled player valuation using data.

  Predictive Analytics Logistic Regression Sports Analytics Python
  Published GitHub ›

• Quantum Approximate Optimization, Traveling Salesman

  Implementation and analysis of the Quantum Approximate Optimization Algorithm applied to the NP-hard Traveling Salesman Problem. Benchmarks quantum approaches against classical solvers and studies where quantum advantage emerges at scale.

  Quantum Computing QAOA Combinatorial Optimization Python
  Published GitHub ›

• Demystifyd

  Helped product-manage a platform built to connect students and professionals internationally as they grow their careers. Contributed to discovery, product strategy, and go-to-market for the web product.

  Product Management MVP Web Platform Go-to-Market
  Live Visit ›

• EZRecruit: Recruitment Management Platform

  A proactive recruitment platform for university varsity coaches that consolidates recruit data, automates updates, and tracks interactions. Built end to end, from discovery through product strategy, design, and financial modeling (Stanford CEE250).

  Product Strategy MVP Web App Financial Modeling
  Product Product Review ›

• Predicting UFC Fight Outcomes

  Machine learning models predicting MMA fight outcomes from fighting style, physical attributes, and experience. Combines logistic regression with a feed-forward neural network using dropout regularization (Stanford CS221).

  Machine Learning Logistic Regression Neural Networks Python
  Research GitHub ›

• Undergraduate Research Archive

  Complete collection of undergraduate projects and academic papers spanning AI, optimization, applied mathematics, and systems. The full record of the work built before Truth Computing.

  Research AI Optimization Applied Mathematics
  Archive GitHub ›

Project and course references describe academic and personal work by Truth Computing’s founders. References to Stanford University (including course numbers such as CS238) describe coursework and research and do not imply that Truth Computing is affiliated with, sponsored by, or endorsed by Stanford University. Product, project, and technology names are used for identification only and remain the trademarks of their respective owners; their use does not imply affiliation or endorsement. Linked repositories are hosted on individual founders’ accounts and may be governed by their own license terms.