How a 90-Minute Prototype Could Return 20 Hours to a County Treasurer

Decision summary

The Lewis County Treasurer's Office had approximately 1,000 parcel records left to review for a municipal annexation project. The existing process required about 20 hours of manual work.

During a 90-minute working session, we mapped the process, built a local automation prototype with Claude, and tested it against records with known results. The work was coordinated with county IT network specialists so testing and development could proceed safely within a limited, managed scope. Initial tests took about five seconds per parcel, producing a projected runtime of approximately 1.5 hours for the remaining batch.

The prototype transferred routine information between the county's property-tax application and an Excel workbook. Missing records, exemptions, and conflicting values were routed to county staff for review.

The projected runtime is based on initial test performance. A completed production run was not documented in the source session.

Business problem

• Workload: approximately 1,300 parcels in the project
• Completed manually: about 300 parcels
• Remaining: roughly 1,000 parcels
• Manual pace: about 100 parcels every two hours
• Estimated remaining effort: approximately 20 hours

The work was structured and repetitive. It required copying parcel identifiers, searching the tax system, reviewing payment information, and updating spreadsheet fields. Most records followed defined rules. A smaller group required staff judgment.

The opportunity was to automate routine records and create a clear review queue for exceptions.

Solution

The Treasurer demonstrated the workflow using live examples. We documented the application steps, spreadsheet destinations, payment rules, and exception conditions.

Claude helped generate and revise a local automation that could:

• read the next parcel identifier from the workbook;
• search the county application;
• retrieve the relevant payment information;
• update the corresponding spreadsheet row;
• preserve supporting values; and
• flag uncertain cases for review.

The prototype supported administrative processing. County employees remained responsible for tax determinations and the final accuracy of the project.

Controls

The implementation included:

• coordination with county IT network specialists for safe testing and development within a limited, managed scope;
• local processing on the approved county computer;
• limited access to the applications required for the task;
• testing against previously completed records;
• explicit handling for missing or conflicting information;
• an internal technical review of generated code; and
• human review of exceptions and final results.

These controls allowed the team to test quickly while preserving operational accountability.

Expected impact

• Implementation session: about 90 minutes
• Observed test speed: about five seconds per parcel
• Projected batch runtime: about 1.5 hours
• Manual baseline: approximately 20 hours
• Expected capacity returned: nearly 20 hours of the Treasurer's time

The office also gained a repeatable process, consistent application of routine rules, and a visible list of records requiring attention.

Why this matters to an executive

Twenty hours represents roughly two and a half working days. Returning that time to an elected official or senior employee increases the organization's capacity without adding staff or replacing its core systems.

For a public agency, the benefit is straightforward: taxpayer-funded expertise can remain focused on financial oversight, public service, and decisions that require accountability. For a private business, the same pattern appears whenever an owner or specialist spends days moving structured information between systems.

Recommended decision criteria for similar projects

Leaders should look for processes with:

1. a measurable manual time requirement;
2. a stable sequence followed across many records;
3. clear examples of normal and exceptional outcomes;
4. known records that can be used for validation;
5. a safe path for human review; and
6. enough recovered capacity to justify implementation and oversight.

This project met those conditions and produced a useful prototype within one working session.

To evaluate a repetitive process in your organization, contact 10 Fold Consulting.