TPRM Last-Mile Automation: SAP-Excel-Outlook

Conceptual framework for an automation system designed to bridge ERP data silos with automated compliance reporting

Project Overview

In Third-Party Risk Management, manual data entry is a significant bottleneck. This project documents a "last-mile" automation workflow designed to bridge the gap between SAP (ERP) and Outlook (Reporting).

Workflow Diagram

graph TD
    subgraph "External Systems (SAP ERP)"
        A[Button Trigger] --> B(VBA Keyword Processing)
        B --> C(SAP Script Execution)
    end
    subgraph "Local Processing (Excel RPA)"
        C --> D(Custom Report Generation)
        D --> E{File Already Open?}
        E -->|Yes| G[Prompt: Close File]
        G --> E
        E -->|No| H(Pricing Logic Check)
    end
    subgraph "Compliance & Output"
        H --> I{Tariff, Prev PO Available?}
        I -->|Yes| J[Apply Price]
        I -->|No| K[Flag for Manual Review]
        J --> L[Final Excel Review]
        K --> L
    end
    subgraph "Mail List & Report"
        L--> M(Outlook mail template draft)
        M--> N{Vendor recorded?}
        N -->|Yes| O[Apply emails receiver]
        N -->|No| P[Flag for Manual Review]
        O --> Q[Final Outlook draft Review]
        P --> Q
    end

Loading

Business Impact

• Efficiency: Reduce 80% of manual workflow filing.
• Accuracy: Eliminated copy-paste errors in sensitive pricing.
• Compliance: Ensured real-time notification for non-agreed rates.

Project Roadmap & Future Enhancements

​This framework is designed for modular growth. Current development is focused on scaling data volume and diversifying input sources:

Phase 1 (Completed): Core ERP-to-Reporting Pipeline. Developed logic for data merging, pricing validation, and automated email drafting with 100% data integrity.

Phase 2 (Completed): DuckDB Integration. Migrating local storage from flat files to DuckDB to enable high-speed analytical queries on multi-year procurement datasets.

Phase 3 (In Development): TF-IDF logic implementation. Improvement of fuzzy match for multi-year procurement datasets.

Phase 4 (Upcoming): OCR Module. Implementing Tesseract-based OCR to ingest digital assessments and quotations into the digital pipeline.

Phase 5 (Exploratory): Local LLM Agent. Testing Ollama/AzureAI integration via REST API for automated document analysis on multi-year sourcing datasets.

---

Note: This repository contains logic frameworks and pseudocode to demonstrate technical proficiency. Proprietary company code is not included.