I design reliable LLM workflows for real-world operations (support agent-assist, KB generation, evaluations) and creative systems (image-to-poetry, storytelling, caption engines). Background in IT support + documentation; building AI assistants with structured outputs, guardrails, and testable prompt iterations.
- Ticket Response Copilot — drafts support-ready replies with troubleshooting steps + escalation criteria
- KB Article Generator — turns solved tickets into consistent knowledge articles (grounded in provided notes)
- Prompt Evaluation Harness — rubrics + regression tests for prompt reliability
- PhotoPoet Storyteller — creative assistant for poetry, captions, and short narratives
Prompt engineering • conversation design • structured outputs (JSON/Markdown) • evaluation & rubrics • safety guardrails • Python • documentation
LinkedIn: https://www.linkedin.com/in/mmartin2025 GitHub: https://github.com/GrandMaTech76