Automation and AI that remove manual work
Built a delivery-automation engine that cleared 703 stuck deliveries in one run and a consolidated data warehouse — plus Copilot/Power Platform agents across operations.
Key outcome
703 cleared in one run · 27+ platforms served
Executive summary
Built a delivery-automation engine that cleared 703 stuck deliveries in one run and a consolidated data warehouse — plus Copilot/Power Platform agents across operations.
Situation
Operations carried a backlog of overdue platform deliveries and ran on spreadsheets and inboxes, with no consolidated view of titles, versions, and storage.
Challenge
The backlog and the manual triage behind it would only grow with the library. Adding headcount would have scaled the cost of the problem, not solved it.
My role
Nick designed and built the automation and the data layer himself, then generalized them across the operation.
Actions
- 01Built automation that cleared 703 overdue deliveries in a single run.
- 02Generalized it into a delivery engine supporting 27+ platforms.
- 03Stood up a consolidated warehouse indexing ~23.7K titles / ~95K versions / ~1.68 PB.
- 04Deployed Copilot Studio and Power Automate agents and authored 213 inbox-automation rules.
Systems involved
Results
- 703 overdue deliveries cleared in one automated run.
- A reusable delivery engine across 27+ platforms.
- A single operational warehouse powering the weekly ops meeting.
Transferable value
Practical AI and automation beat headcount when the tools are observable and maintainable. The win is removing manual work from the critical path, not chasing novelty.
Related skills