This repo contains work samples in AI operations, workflow QA, data integrity, and system design - built to keep outputs consistent and workflows traceable.
- Chargeback Defender - e-commerce dispute operations kit: Systems/Chargeback-Defender
- HubSpot CRM - messy data import and cleanup case study: HubSpot-CRM-Data-Cleanup-Case-Study.md
- Relocation Roadmaps - project summary (PDF preview): Relocation-Roadmaps-Project-Summary.pdf
- AI Guide Audit Workflow - end-to-end overview: Systems/AI-Guide-Audit-Workflow
- Run AI-assisted workflows with real-world inputs without losing quality or introducing drift
- Build prompt templates, SOPs, and QA checks that keep outputs consistent and reduce rework
- Keep data usable through standardization, deduplication, required fields, and naming conventions
- Troubleshoot operational issues, identify root causes, and document fixes so problems do not repeat
- Design practical workflow systems and carry them through to completion
The Relocation Roadmaps samples come from a real project - Relocation Roadmaps - and are shared to show process, structure, and decision-making, not as polished client deliverables.
The Chargeback Defender samples are from a live product built and shipped in 9 days - Chargeback Defender.
- Chargeback Defender artifacts: decision calculator, dispute tracker with rate monitoring, evidence assembly guide, test cases, project conventions, and schema documentation - Systems/Chargeback-Defender
- Relocation Roadmaps artifacts: project summary, reviewed guide sections, evaluation scorecards, workflow notes, and redacted prompt excerpts showing how structured output was produced and checked - Systems/Relocation-Roadmaps
- AI audit workflow artifacts: structured outputs (TSV and JSON) and review tracking across Sheets, Zapier, and Notion
- QA artifacts: test plans, test cases, and sample bug reports
- AI workflow examples: evaluation scorecards and reviewed guide sections
- Prompt system excerpts: redacted prompt templates used to produce structured output
- Systems notes: brief documentation of live automations
