Machine learning systems
Built and deployed computer-vision applications with optimized model formats, REST APIs, user interfaces, generated documentation, and CI-driven artifacts.
AI consulting · automation · AWS
Practical AI systems for real workflows.
I help teams take the workflow they already have and turn it into an AI system that survives production — LLM-powered tools, automation, data pipelines, and deployment on AWS.
engagement model
Three phases. Enter where it hurts.
-
AI opportunity audit
Map your workflows, data constraints, risks, and likely ROI into a practical implementation plan.
a plan you can execute — with or without me
-
Prototype sprint
Build a focused proof of concept: RAG assistant, document workflow, automation tool, evaluation harness, or data interface.
working software plus honest eval results
-
Production implementation
Deploy an AWS-hosted service with APIs, logging, CI/CD, cost controls, documentation, and handoff.
a system your team owns and can run
relevant work
Built for production, not demos.
CV / ML
DATA
Data pipelines and model readiness
Developed pipelines that turn raw sensor data into training and evaluation datasets — annotation workflows, class-imbalance handling, experiment tracking, and validation baselines.
AWS
Cloud and automation
Advised on AWS data infrastructure, automated deployments, supported ETL into relational stores, and translated technical progress into sponsor-ready milestones and decisions.
EDGE
Edge and constrained deployment
Deployed models in ONNX and TensorRT formats, supported Nvidia Jetson workflows, and worked across autonomy, simulation, sensor fusion, and fielded robotic systems.
about
Engineer first, consultant second.
I am a software and machine-learning engineer with experience across autonomous systems, sensor-data pipelines, ML deployment, AWS infrastructure, proposal evaluation, and cross-functional technical leadership.
M.S. in Computer Science, Georgia Tech. B.S. in Computer Science, Florida State University. Security+, Network+, and A+ certified.
contact
Have a workflow that should be a system?
Send me four lines. That's enough to tell you whether this is worth pursuing — and roughly what it would take.
# brief.md — what to send workflow: what happens today, step by step data: what exists and where it lives users: who touches it, and how often success: what changes if this worksSend it on LinkedIn →