Public builds

Public projects, personal interest pieces, and ongoing research and development.

A working collection of small, complete builds — some have ended up in production, others are experiments and prototypes I keep returning to.

M5 Forecasting Accuracy
Public build Forecasting benchmark

M5 Forecasting Accuracy

End-to-end global ML forecasting solution on a well-known business problem: 28-day daily forecasts across 30,490 Walmart product–store series, with three model families (Theta, AutoETS, LightGBM) and rolling-origin cross-validation.

Approaches world-class accuracy with a deliberately small feature menu — built around reproducibility, deterministic runs, and clarity rather than leaderboard chasing.

Python Nixtla LightGBM Polars Terraform AWS Azure GCP Hetzner
FFPy — Fantasy Football Optimization Engine
Public build Open source toolkit

FFPy — Fantasy Football Optimization Engine

A Streamlit app and Python toolkit that ingests nflverse and ESPN data into a local SQLite store, generates weekly projections, and optimizes lineups via linear programming.

An end-to-end build — data ingestion, projection modeling, constrained optimization, and a usable interface — in a domain where decision quality is measurable every week.

Python Streamlit PuLP nflverse
BASB — Building a Second Brain
Research & development Personal knowledge OS

BASB — Building a Second Brain

A Tiago Forte–inspired personal knowledge and tooling system: a one-shot installer that bootstraps a fresh machine for daily ML/AI work, plus a small CLI (arkolab) for GitHub profile and workflow automation.

Captures notes, tools, and workflows in one place so context follows me across machines instead of living in scattered files.

Python Bash uv Automation
PiCLI — Raspberry Pi Secure Setup Toolkit
HomeLab Device hardening toolkit

PiCLI — Raspberry Pi Secure Setup Toolkit

A command-line setup toolkit for turning a Raspberry Pi into a safer, repeatable Linux host with secure defaults and automation-first setup steps.

Codifies the first-mile hardening work that is easy to forget on small devices: repeatable bootstrap, safer access patterns, and a setup path that can be rerun instead of reconstructed by hand.

Bash Raspberry Pi Linux SSH Security Hardening Automation
gifCLI — Recording-To-GIF Release Tooling
Public build Developer tooling

gifCLI — Recording-To-GIF Release Tooling

Convert video (or any ffmpeg-supported input) to optimized GIFs.

Makes release communication faster by packaging terminal or demo recordings into embed-ready GIFs with repeatable commands.

Bash ffmpeg CLI Automation
OpenCharles — Secure OpenClaw Multi-Model Harness
HomeLab Home AI lab

OpenCharles — Secure OpenClaw Multi-Model Harness

A Raspberry Pi 5 setup for running OpenClaw in a secure home-lab environment, with a multi-model harness for local AI experimentation.

Makes edge AI lab work more reproducible by treating model routing, host setup, and security posture as infrastructure instead of one-off terminal history.

Raspberry Pi 5 OpenClaw Linux Multi-Model LLMs Home Lab Security
LG
HomeLab Creative automation tool

Logos — AI Project Logo Generator

An AI-assisted project logo generator for producing lightweight visual marks that can travel with small builds, labs, and experiments.

Turns project branding into a repeatable workflow, giving prototypes a recognizable visual identity without dragging a small engineering project into a full design process.

AI Image Generation Prompting Design Automation Project Branding
GoodReads
Public build Recommendation prototype

GoodReads

Learning project using good reads data.

A simple public artifact that mirrors my longer-running interest in recommendation and decision support systems.

Python Recommendation Systems Collaborative Filtering Content-Based Filtering Polars Spark
Public build Local-first toolkit

CareerOS

A local-first, fork-friendly career operating system.

Treats career artifacts — applications, networking, deliverables — as data you own, version, and rerun like any other engineering workflow.

Python Typer Flask Jinja2
VN
Public build Optimization research

VN1 Inventory Optimization

Open ML / MILP solutions for the VN1 inventory optimization problem.

Pairs probabilistic forecasting with mixed-integer programming to expose how model accuracy translates into real inventory decisions.

Python MILP PuLP Forecasting