Rodney E.

An AI agent can finish a task. I build and monitor the workflow around it—inputs, approvals, handoffs, and what happens when something breaks. For founders and small teams, that can mean turning incoming information into organized tasks, research briefs, documentation, and recurring reports. My main workspace is Hermes Agent. I use Codex for coding and implementation. ━━━━━━━━━━━━━━━━━━ WHAT I BUILD → Agentic Operations Task intake • Routing and delegation • Scheduled checks • Human approvals • Exception monitoring • Status updates → Research & Reporting Systems Source tracking • Content classification • Airtable records • Structured briefs • Recurring reports • Traceable inputs → Startup Documentation Process maps • SOPs • Decision logs • Requirements • Handoff notes • Workflows people can maintain → Product UX Strategy User journeys • Onboarding reviews • Permission and system-state clarity • Flow prototypes • Provisional design-system direction → AI-Generated Visual Content Photo concepts • Video concepts • Iteration and selection • Human quality review using fal, Higgsfield, and MiniMax ━━━━━━━━━━━━━━━━━━ HOW I WORK I start with the work, not the agent. • What comes in, and where does it come from? • What needs a decision—and who should make it? • What should happen when a step fails or the information is uncertain? • What does a useful finished result look like? Then I build a bounded first version, test it, document the handoff, and monitor how it performs. I keep a person in the loop where the decision or outcome needs one. ━━━━━━━━━━━━━━━━━━ MY STACK Agentic workflows Hermes Agent (primary) • Codex (coding and implementation) Data & operations Airtable • Notion • Google Workspace • CSV/JSONL workflows Product & visual work Figma • fal • Higgsfield • MiniMax ━━━━━━━━━━━━━━━━━━ SERVICES Agentic Workflow Design • AI Operations Monitoring • Research & Reporting Systems • Startup Documentation • Product UX Strategy • AI-Generated Photo & Video Concepts ━━━━━━━━━━━━━━━━━━ HAVE A WORKFLOW IN MIND? Send me the steps you currently follow, a sample input, and the result you want. I'll help identify what can be automated, what needs human review, and what a practical first build would look like.