Case Studies  ·  April 08, 2026

AI Document Intelligence Cuts Chaos on a $2M+ Construction Project

The Problem

Moneybrook Farm LLC was managing a $2M+ construction project with over 5,000 documents scattered across email, Google Drive, and paper — with no search, no organization, and no reliable way to surface critical dates or action items. Institutional knowledge lived entirely in one person's head, creating a single point of failure for the entire project.

What We Built

We built an AI-powered document intelligence system that automatically classifies, names, versions, and files incoming documents with 90%+ confidence, while continuously monitoring Gmail every 12 hours for new material. A vector database enables natural language search across the entire document library, and an action items dashboard tracks 24 items automatically — giving the client full visibility without manual effort.

How It Works

1. Step 1: Gmail API monitors the client's inbox on a 12-hour cycle, pulling in new construction-related emails and attachments automatically.
2. Step 2: Claude API classifies each document with 90%+ confidence, then the system applies professional naming conventions and versioning before syncing files to Google Drive.
3. Step 3: Documents are indexed into a ChromaDB vector database, enabling natural language Q&A so the client can ask plain-English questions and instantly retrieve critical dates, obligations, or action items from across 5,000+ files.
4. Step 4: An action items dashboard aggregates and tracks 24 items automatically, surfacing what needs attention without any manual triage.

Tools Used