Approved-tool limits
The source systems were difficult to automate with the standard approved desktop automation options.
Case Study 01 • Reporting Automation
A week-long, distributed billing-report cycle for New York City government accounts became a repeatable workflow that one analyst could manage—while preserving review points for exceptions.
Overview
During my T-Mobile reporting role, the monthly process supported multiple New York City government accounts and a billing arrangement in which different charge categories required separate handling. Deliverables were segmented by business unit, project, account structure, and service-line constraints.
The work was not simply file conversion. Each Excel/PDF pair had to represent the same account-level detail, survive changing source conditions, and remain reviewable before delivery.
Problem
The original week-long workflow spread source retrieval and report creation across 5–7 team members. Analysts opened files, copied details into an invoice-style workbook, removed unused rows, exported a PDF, and paired the two deliverables. Repeating that sequence across 150+ account-based packages made consistency depend on careful manual execution.
The source systems were difficult to automate with the standard approved desktop automation options.
Packages varied by business unit, project, account number, and line-count rules rather than one simple customer split.
Service-code changes and source gaps could create mismatches between line items and totals.
Excel and PDF versions had to stay aligned across a large set of recurring deliverables.
Approach
I separated the workflow into stages so that retrieval, processing, output generation, and review each had a clear responsibility. This reduced repeated interaction with individual files while keeping a deliberate checkpoint for source issues.
A JavaScript-assisted browser workflow looped through accounts and downloaded the recurring source files.
Downloaded files were organized in a predictable folder structure for the reporting run.
Excel VBA processed the folder, applied the template logic, and produced the final Excel and PDF package pair.
Checks compared line items with totals and flagged mismatches for manual research before delivery.
Sanitized Evidence
Source exports move through controlled staging and shared rules before the same run creates the Excel and PDF pair. Mismatches leave the main path for analyst review.
Key Decisions
Reasoning
The source environment did not expose a straightforward supported bulk-export path. A focused script reduced repetitive account-by-account work without pretending the entire upstream system was under my control.
Tradeoff
The retrieval step depended on the source interface and still required an informed operator.
Reasoning
Creating the Excel and PDF outputs together reduced the risk that paired deliverables reflected different edits or source states.
Tradeoff
The template logic had to handle formatting and file naming as carefully as the data itself.
Reasoning
When source gaps or service-code changes produced a mismatch, a visible review path was safer than silently normalizing the discrepancy.
Tradeoff
Some cases remained manual by design, but the analyst investigated only the flagged exceptions.
Tools & Validation
Result & Impact
The automation reduced the number of manual handoffs, improved consistency between paired outputs, and created capacity for other analysts. Just as important, validation was part of the workflow rather than an informal step performed after generation.
Learnings
This case study is intentionally sanitized. Individual agency names, account structures, and source screens are omitted, while the public NYC context, operational scale, and decision logic are preserved. The result should be read as a reporting-process improvement within the tools and permissions available—not as replacement of the upstream billing systems.