Skills
What I build with and maintain
Programming Languages
Frameworks
Tools
Experience
Professional experience and education
Jan 2026 — Present
Systems Engineer
ImageNet (MSP)
Architected Azure AD and Microsoft 365 tenant configurations, managing 500+ user identities across hybrid cloud environments. Automated endpoint provisioning workflows using PowerShell scripts, reducing deployment time by 60%.
Dec 2025
B.S. Computer Science
Florida Atlantic University
Graduated from FAU with a Bachelor of Science in Computer Science. Built production-grade web applications and automation tooling throughout the program.
Feb 2024 — Jan 2026
Technician
Treasure Coast IT Solutions (MSP)
Designed and implemented network infrastructure solutions including VLAN segmentation and secure VPN access for multi-building facilities. Managed IT infrastructure for 30+ sites, overseeing Windows/Linux servers, networking, and security systems. Guided a 3-month university student internship leading a hands-on automation project.
Projects
Things I've built
Distributed AI Agent Platform for Autonomous Open-Source PR Fixes
Distributed AI agent platform for autonomous open-source PR fixes. Claude Code worker fleet with atomic Postgres job dispatch (UPDATE...RETURNING) eliminates the message-broker dependency, enforces per-user token budgets, and uses graceful SIGTERM checkpointing with WIP commit recovery so handoff workers resume from a real recovery point. A custom MCP stdio server exposes 8 purpose-built tools (queue claim, git autocommit, incremental token reporting, WIP handoff, structured escalation with machine-readable reason codes) as the build agent's sole control plane. The 5-stage GitHub issue pipeline — trending scan → Haiku LLM fixability classifier → human review gate → fork-PR dispatch → PR state polling — runs through a 7-state fix-candidate machine with pre-dispatch safety filters. Embedding-based pain-signal clustering across Reddit, Discord, and GitHub Issues uses cosine similarity over a rolling 30-day window to deduplicate complaints into ranked priority signals. Production FastAPI service: 15 route modules, 30 Alembic migrations, daily backups, cron harvests processing 50 trending repos and 100+ classify jobs per tick.
Open Source Contributions
PRs shipped to repos I don't maintain
~600,000x speedup on the Resize operator via separable interpolation
ONNX (Open Neural Network Exchange) is the Linux Foundation AI standard for ML model interoperability, used by PyTorch, TensorFlow, scikit-learn, and every major ML framework. Rewrote the ReferenceEvaluator Resize operator using separable 1-D interpolation, replacing per-output-pixel recursion with vectorized numpy operations. Achieved ~600,000x speedup (15 minutes to 1.5 ms on a typical (1,32,20,20) → (1,32,40,40) bilinear/half_pixel resize) while preserving bit-exact numerical compatibility (max diff <1e-10) across 294 test configurations covering every mode × antialias × coordinate_transformation_mode × shape combination, plus tf_crop_and_resize extrapolation and axes-restricted resizes. Resolved a long-standing performance issue (#6554) open since November 2024. Full 1798-test backend suite passes; ruff and mypy clean.