Case Studies

What I ran, and what changed

Internal documents and plans stay confidential. These describe what I did, at what scale, and what changed as a result.

Running a SaaS product with AI agents in the loop

NeatDrive · Google Drive sharing-audit tool · solo founder · 2026–present

Situation. One person, a production codebase, and a product that had to launch, pass marketplace review and serve real users without a team.

What I did. Put AI agents (Claude, Claude Code) on the build, audit and reporting work, and wrote the operating system around them: dispatch protocols that say how a task is scoped and reported, a three-part test gate every change passes before it ships, a failure taxonomy of the ways agent output goes wrong, and a decision register and defect register so the operation stays auditable. When an agent’s report asserted something, the rule was to verify it against the artifact before acting on it.

What changed. Launched August 2026; listed on the Google Workspace Marketplace after review. The verify-before-acting rule caught agent reports that asserted things the code didn’t support. This is what “practical fluency in AI-enabled operations” looks like when it is your own production system on the line.

A benefits decision made on the numbers

MiaDonna & Company · 10–20 employees · multi-state

Situation. Benefits renewal with quotes in hand, a broker relationship by habit, and no view of what the quoted plans would actually cost staff against how they enrolled.

What I did. Modeled the quoted plans against the team’s actual elections, and put the comparison in front of the owner with a recommendation.

What changed. The company changed brokers, and the owner moved other teams to the same broker afterward.

Finance operations: $60K a year back, and 23 working days

MiaDonna & Company · multi-million-dollar budgets · 2021–2026

Situation. Vendor contracts and SaaS subscriptions nobody had audited; accounts receivable running on manual entry; a chart of accounts that made reporting slow and cash flow hard to see.

What I did. Audited every vendor and SaaS contract and renegotiated, replaced or cut. Built an automated AR workflow. Redesigned the chart of accounts to GAAP.

What changed. $60,520 in annual recurring savings, a 40% cut in recurring expenses. The equivalent of 23 working days a year returned to staff. Leadership and lenders working from the same numbers, faster.

A ~$200K flagship store relocation

MiaDonna & Company · Bridgeport Village

Situation. Flagship retail store to relocate and open, with construction running late.

What I did. Lease negotiation and compliance, vendor procurement, construction logistics, initial inventory onboarding, and the trade-off calls as the delays landed.

What changed. Opened on schedule, on a ~$200K budget.

Payroll and policy across five states and Canada

MiaDonna & Company · 2021–2026

Situation. A company going remote-first, hiring across state lines, with a handbook written for one office.

What I did. Ran payroll and compliance end to end across five states and Canada; rewrote the handbook for sick leave, safety and remote work; administered BambooHR as the HRIS; handled I-9s, workers’ compensation and unemployment claims.

What changed. A multi-state payroll operation and a policy framework the company could hire into.

One workbook for a 20,000-person conference

ISTE · international education nonprofit · 2018–2022

Situation. 12,500–20,000+ participants and up to 3,000 presenters a year, with presenter, volunteer, staff travel and hotel assignments tracked across dozens of separate documents.

What I did. Replaced them with one automated workbook, and analyzed 2,000+ post-conference feedback surveys so the team had a measured view of what worked.

What changed. Duplicate entry and errors cut; leadership had a live view of logistics; the next year’s event was planned from data rather than recollection.

Moving a statewide team onto one intranet during COVID

Oregon Primary Care Association · 2020

Situation. The pandemic sent staff home from an organization running on local servers.

What I did. Migrated the organization to Microsoft Teams and SharePoint, built the central intranet, ran the RFP for a new IT managed service provider, and wrote a 40-page Remote Work Manual.

What changed. A statewide team working remotely within the window, on documented procedures rather than tribal knowledge.