About Experience Projects The Lab Writing Let's Talk

Hi, my name is

Jay Stuart.

I build what matters.

AI & cloud consultant. I build generative AI, LLM, and multi-agent systems — and the cloud platforms under them — for startups, enterprise teams, and federal programs where getting it wrong isn't an option.

26 years of building systems — from COBOL to Kubernetes to generative AI and multi-agent frameworks. Work spans startups, growth-stage SaaS, enterprise, and federal agencies (NOAA, HHS, FDA, VA, CMS, IRS).

Currently focused on applied AILLM products, RAG platforms, AI agent systems, and the cloud infrastructure under them. Pioneered AI agent frameworks delivering 6x acceleration; architected cloud platforms under FedRAMP/NIST 800-53 and contributed to $50M+ in contracts along the way.

Builds in the open, runs a production K8s homelab, cooks seriously.

Available for remote AI and cloud consulting anywhere in the US — generative AI, LLMs, multi-agent systems, RAG, cloud architecture, DevOps, and fractional CTO/AI leadership. Based in Baltimore, on-site across Washington DC, Northern Virginia, and Philadelphia. Private sector, startups, and enterprise welcome.

"Everything's been done before. My job is to make it feel like the first time."

Generative AI LLMs RAG AI Agents Multi-Agent Systems LangGraph OpenAI / Claude AWS / Cloud Kubernetes Python DevOps Terraform FedRAMP

Principal Solutions Architect

SAIC
Stabilized NOAA's AWS cloud infrastructure and built governance frameworks from scratch. Designed future-state architecture for mission-critical weather and climate systems.
— Present

Chief Architect

Monarch Innovations
Pioneered AI agent frameworks achieving 6x delivery acceleration. Leading architecture for intelligent automation, multi-agent governance, and self-healing test systems.

Senior Lead Technologist

Booz Allen Hamilton
Contributed to $50M+ in contract wins. Delivered FedRAMP/ATO-compliant cloud solutions for HHS, FDA, and VA. Led teams of 20+ engineers across multiple federal programs.

Technical Lead

ActioNet
Built knowledge management platforms for CMS. Streamlined content workflows and collaboration tools for federal healthcare programs.

Principal Architect

AT&T
Led enterprise Java architecture and drove Agile transformation across engineering teams. Built platforms serving millions of subscribers.

VisionTest AI

visiontest.ai

AI-powered visual regression testing that catches what unit tests miss. Automated screenshot comparison with intelligent diff detection.

AI/ML Computer Vision Testing

SourceBridge AI

sourcebridge.ai

Codebase intelligence platform that understands your entire repository. Semantic search, dependency mapping, and AI-driven code analysis at scale.

RAG LLMs Code Analysis

Project Athena

local • apple silicon

Fully local AI voice assistant running on Apple Silicon. No cloud dependencies — private, fast, and running on a Mac Studio cluster in the homelab.

Local LLM Voice AI Apple Silicon

AgentPulse

real-time agent observability

Live dashboard for Claude Code and Codex CLI sessions. See what every agent is doing across your terminal tabs — prompts, tool usage, and session history — with orchestration for launching and managing runs.

Developer Tools Observability AI Agents

Mozart Orchestration

claude code plugin

Turns one request into a narrated, multi-agent delivery pipeline — research, plan, specialist review, implement, verify, document. Named subagents run in their own contexts with handoffs you can watch in real time.

Claude Code Multi-Agent Orchestration

A personal platform for experimenting with distributed systems, local AI inference, and resilient infrastructure. Not a collection of hardware — a proving ground for architectural decisions before they reach production. 10 GbE core network, 120 Gbps Ceph backend, separate dev environment for staging workloads.

4-node
Proxmox cluster · 96 GB RAM each
448 GB
Cluster RAM (384 GB compute + 64 GB dev node)
~5 TB
Ceph distributed storage, 3× replicated (16 TiB raw)
22 TB
Synology NAS — backups & persistent volumes
208 GB
Apple Silicon unified memory for AI inference (128 + 64 + 16 GB)
30+
Self-hosted services running continuously
Runs → Project Athena (local AI voice assistant) · Qwen 3.5 35B MoE inference (powers the chat widget on this site) · SSO, monitoring, docs, dev staging, and more — all private, all on-prem
GitHub Activity
Loading activity...

Let's Build Something.

Now taking on AI & cloud consulting engagements — generative AI strategy, LLM integration, RAG platforms, AI agents and multi-agent systems, AI product engineering, and the cloud / Kubernetes architecture under them. Also available as a fractional CTO, AI leader, or Chief Architect, and for Principal / Chief Architect and VP Engineering roles.

Private sector first — startups, growth-stage SaaS, and enterprise teams. Federal clients welcome too. Fully remote anywhere in the US, with on-site availability from Baltimore, MD across Washington DC, Northern Virginia, Philadelphia, and the broader Mid-Atlantic.

Say Hello
Ask Athena anything
Project Athena Local AI
Running on Jay's homelab · Qwen 3.5 35B
Hi! I'm Athena — Jay's personal AI assistant, running locally on his homelab (no cloud APIs). Ask me about his experience, projects, or what he's building. What would you like to know?