Replacing Excel with SAP APIs & Python validation

Many enterprise reportings still depend on manual Excel uploads, fragile scripts, and “don’t-touch-it” legacy code. That mix makes reporting slow, error-prone, and hard to audit — especially with increasing governance expectations.

In this session, I share a practical modernization: replacing Excel-based inputs with direct SAP API extraction, transforming the data in Python, and integrating the output into an existing reporting solution. The work started with understanding undocumented legacy scripts and the implicit business rules they encoded; AI-assisted developer tools (e.g., Copilot) helped to accelerate code understanding before making changes.

Two aspects shaped the final design. First, API-derived data deviated from historical reports because of synchronization/timing effects — reports had to be generated in the correct time window to be consistent. Second, replacing the manual Excel layer surfaced a long-standing deviation between the intended business logic and the legacy report’s actual behavior. This created an opportunity to streamline and harmonize the logic, improving consistency, runtime, and scalability of the reporting pipeline while reducing manual effort and key-person dependencies.

To rebuild trust, a validation harness was create: a Python script that automatically compares the “manual” and “automated” outputs, flags deviations, and visualizes differences so reviewers can triage issues quickly. This reduces reliance on individual availability (vacation, sick leave) for critical reporting checks and made the whole process more resilient. Unlike generic “automation with AI” talks, this is a governance-first story about validation, rollout, and preserving reporting integrity in production. You’ll leave with concrete patterns for baselining, reconciliation, and monitoring when modernizing enterprise pipelines.