Featured work // Public code // Private vault
Open source contributions, production systems, and side projects
High-performance time-series forecasting library written in Rust. A reimplementation of Facebook Prophet optimized for production workloads. Currently running in production for mission-critical decision systems handling billions in freight pricing.
View Project →Lightweight NLP recommendation engine using Gemma 270M and cosine similarity for personalized product matching. Demonstrates efficient deployment of small language models for specialized domains.
Built enterprise-scale agentic AI platform using LangGraph/LangSmith that processes $80B+ in annual freight bids. Automatically extracts and processes information from PDFs, emails, and Excel files. Reduced turnaround time by 40 hours and freed up 20 FTEs.
Created Model Context Protocol (MCP) server infrastructure enabling agent systems to pull real-time costing guidance and deliver consistent decision support to commercial representatives. Integrates with existing data warehouses and pricing engines.
Deployed conversational AI agents that provide on-demand pricing intelligence and recommendations. Enables business users to query complex data warehouses using natural language and receive context-aware insights.
Built webhook-based automation platform that captured real-time events and automated task management across IBM Finance departments. Resulted in time savings equivalent to over 50 FTE. Integrated with Trello, Slack, and internal IBM systems.
A July 2026 snapshot of 33 owned public repositories, followed by one privacy-preserving panel representing 25 qualifying private repositories. Public entries are sorted by most recently pushed.
Check out my GitHub profile for more open source work and experiments.
Compare lump-sum investing and DCA across real SPY history.
An interactive homage to Jorge Luis Borges's The Library of Babel.
A skill and web interface for turning strength goals, timelines, and equipment into practical coaching recommendations.
H3 map of registered offenders
ATE is a scam
Simple comparison for when local llms breakeven for cloud
Home maintenance tracking app with FastAPI, React, PostgreSQL, multi-agent support, and Telegram notifications.
Time-series forecasting in Rust based on prophet
AI-assisted Pokémon card collection manager with card recognition, market pricing, and portfolio tracking.
Geospatial Time Series
Explores nearest-neighbor weighting to interpolate low-resolution spatial data.
Scrapes local businesses to surface the best coffee deals.
Telegram bot that stores notes and appointments, adds context, and prepares a daily report.
aoc 2024 code
multi-agent researcher to do multimodal research on companies
I dont have time to read the bills the house passes, lets have an AI do it.
Interactive rocket simulation served as a lightweight local web app.
Plex Server
One repo to rule all of them home things
Repo for the autobsts package
Repo contains a working example of container orchestration in kubernetes to utilize seperate containers for pulling data, modelling data, and pushing data out.
System of ODE's for epidemic modelling put into a R-shiny Dashboard
Explores Chicago crime arrest prediction with caret and SparklyR workflows.
Template Repo for R projects using docker
Containerized model-build template for packaging R workflows for scheduled jobs, APIs, or cloud deployment.
Time Series Package in R to automate forecasting
Explores logistic models for predicting weightlifting medal outcomes from weight class and country.
Looking at the death Penalty
shiny app to look at bmr and how long it takes for people to lose weight
Showing you how to deploy a plumber api on kubernetes in ibm cloud
Messing around with Text Analytics in Twitter
Find your distance relative to Fazoils
R project on looking at world of warcraft gold prices
No represented projects match this search.
Production experience deploying multi-agent systems, RAG pipelines, and conversational AI at enterprise scale
GCP, AWS, Kubernetes, Helm, Docker - building and maintaining production ML systems
Python, Rust, R, SQL - choosing the right tool for performance, maintainability, and team expertise
Snowflake, modern data warehouses, building scalable data pipelines for ML systems