AI Document Parser for Spreadsheet Automation

Customer: AI | Published: 17.10.2025

We’re looking for a skilled automation developer who can design a clean, reliable, and well-documented process that uses AI to extract data from solicitation documents (PDFs or DOCXs) and populate our sourcing spreadsheets in SharePoint. We don’t care what stack you use — Power Automate, Azure Logic Apps, or a lightweight Python or Node-based service — as long as it’s clean, maintainable, and works flawlessly inside our Microsoft 365 environment. Context Video: https://www.loom.com/share/08106474ceb349709f448d6f59f5ac7b?sid=b0f1e00f-eef2-4a8c-b55f-91f4ffe4bffe We operate two connected businesses: • Southern Executive Transport (SET): Transportation brokerage • Apex Procurement Solutions (APS): Federal project management and procurement firm Both teams currently review solicitation documents manually to identify project title, scope, dates, and submission requirements before entering them into a sourcing workbook. Each one takes 15–20 minutes. We want to automate this process so an AI parser can extract key details and write them into the sourcing sheet automatically. Folder Structure Same logic across Federal and SLED projects. Example: Documents → 02_Opportunities → [Federal or SLED] → 01_Active → [Agency or State] → [Project Name] → 01_Solicitation Documents → 02_Opportunities → [Federal or SLED] → 01_Active → [Agency or State] → [Project Name] → 02_Vendor Information → 01_Sourcing & Outreach → APS - Master Sourcing Template.xlsx Automation should detect new files added to 01_Solicitation, then update the matching APS – Master Sourcing Template.xlsx in the same project folder. Core Deliverables 1. Detect new solicitation PDFs/DOCXs in SharePoint folders. 2. Extract text (OCR if needed). 3. Parse and return structured data (using OpenAI, Azure OpenAI, or equivalent). 4. Write results to Excel fields by label (not by cell reference). 5. Flag low-confidence results for manual review. 6. (Optional) Send Teams alerts when a document is processed. Fields to capture: • Project Title • Solicitation Number • NAICS or PSC Code • Due Date / Time • Location • Scope Summary • Line Items (if present) • Submission Requirements • Source Link • File Link Requirements • Experience with SharePoint + Excel automation (any method). • Experience integrating AI parsing (OpenAI, Azure OpenAI, Claude, etc.). • Clean, maintainable code or workflow structure. • Documentation and handoff instructions required. Goal Reduce sourcing prep time from ~20 minutes to under 2 minutes per solicitation. Build a scalable, dependable system that we can later expand to include Prime Outreach automation in 20–30 days.