Tyreeq Harvey — AI / Agentic Workflow Engineer

01 / Projects

Featured projects

Agentic AI · Local LLM Tool-Calling

In Progress / Phase 1

Local Email Agent

A personal email assistant that runs entirely on your own machine. It talks to a local LLM via Ollama in a terminal chat loop, using tool-calling to read your Gmail inbox, compose and send emails, and manage a saved-contacts address book. Nothing leaves your machine except the Gmail API calls themselves — the model runs fully local. Sending is deliberately split into a prepare-and-preview step and a separate explicit-confirm step, so a real send can never happen by accident.

Python
Ollama (Llama 3.1)
SQLite
Gmail IMAP/SMTP

View code on GitHub Watch demo

Java · AWT/Swing → JS/Canvas port

Complete

BLOX — A Tetris Game Built in Java

A full Tetris implementation written in Java: a 10×20 board tracked with a 2D array, seven tetromino shapes, a hand-rolled 90° rotation matrix, wall/floor/collision detection, line-clear scoring, and a Swing-rendered UI with a live score sidebar.

Java
Swing / AWT
2D Arrays
OOP
Game Loop

View Java source

Python · Raspberry Pi · OpenCV / Tesseract OCR

Complete

AutoEntry100 — Smart Garage Access System

A license-plate-triggered garage system built with a team of two — I owned the software stack. A Pi camera feeds frames to an OpenCV/Tesseract OCR pipeline that reads the plate, an ultrasonic sensor confirms the car is in position, and a match against the authorized plate list drives a servo-controlled door, an I2C LCD status readout, and red/green access LEDs — all wired up through a Tkinter GUI for manual scans and live camera preview.

Car pulls up → camera reads the plate → match check → garage opens

Python
OpenCV
Tesseract OCR
Raspberry Pi / GPIO
Servo Control
Tkinter GUI

View Python source

Simulated walkthrough below — the original runs on physical Raspberry Pi hardware (camera, servo, ultrasonic sensor, LCD), which can't run in a browser.

02 / About Me

I'm an aspiring AI engineer focused on building tools and products that solve real problems. I'm always looking to expand my skills whether that is building engineering projects, games, simulations, and tools I personally use to boost productivity. To me, technology isn't about chasing the latest trends; it's about giving engineers the chance to creatively solve real-world challenges.

My background spans backend engineering and applied machine learning, which naturally evolved into AI engineering. I treat LLMs as components in a larger system: observable, testable, and replaceable.

Ultimately, my aim is to bring ideas to life using AI.

03 / Contact

Let's build something

Open to full-time AI engineering roles and contract work on RAG systems, agentic workflows, and model fine-tuning. The fastest way to reach me is the form — or find me here:

github.com/tyreeqharvey linkedin.com/in/tyreeqharvey Download resume (PDF)