Agentic AI &
Multi-Agent Systems
Modern AI applications can do much more than generate text. Agentic systems combine AI models with tools, workflows, application logic, and specialized agents that collaborate on complex tasks within defined permissions and approval boundaries.
Mike has architected a Codex-based multi-agent engineering system with specialized development, security, QA, review, deployment, and orchestration agents, including parallel review, autonomous rework, human approval gates, and integration with Jira, GitHub, and GitHub Actions CI/CD.