Introduction
PIP AI is an enterprise-grade construction pre-construction platform that matches Punch Item Plans (PIPs) and Architectural Floor Plans to corresponding specifications using AI-powered semantic matching.
What is PIP AI?
In construction pre-construction workflows, estimators must manually match hundreds of line items from Punch Item Plans to the correct product specifications. This is time-consuming, error-prone, and expensive.
PIP AI automates this process by:
- Parsing PIP documents using AI to extract structured line items
- Managing a spec library with vector embeddings for semantic search
- Auto-matching PIP items to specs using hybrid search (semantic + keyword)
- Providing spatial context through interactive floor plan overlays
- Generating documents via a drag-and-drop document builder with AI image generation
Core Value Propositions
| Value | Description |
|---|---|
| Zero Mismatches | Every PIP line item links to the exact specification required, eliminating costly spec gaps |
| Spatial Context | Tasks are grounded to physical space — "Renovate Guest Bathrooms" highlights all guest bathrooms on the floor plan |
| Continuous Learning | The system learns from estimator feedback to improve match accuracy across future projects |
| Collaborative Workflow | Centralized hub for the entire pre-construction team to review, annotate, and approve scope |
Target Users
- Senior Estimators — High-precision matching and automated quantity take-offs
- Project Executives — Dashboards, version control, and compliance assurance
- Design/Visuals Team — Brand standard matching and design intent validation
Tech Stack
PIP AI is built with:
- Nuxt 4 (Vue 3 Composition API) — Frontend framework
- Supabase — Backend (PostgreSQL + pgvector, Auth, Storage, Realtime)
- N8N — Workflow orchestration for document processing
- Fabric.js — Interactive floor plan canvas
- Pinia — State management
- TanStack Table — Data tables
Project Workflow
The core application follows a 5-tab sequential workflow per project:
- Upload PIPs — Ingest PIP PDF documents for AI extraction
- Upload Floor Plans — Add spatial context with floor plan images
- Spec Matching — AI-assisted matching with confidence scoring and human review
- Document Builder — Drag-and-drop document creation with AI image generation
- Finalize/Review — Export and final review
Next Steps
- Installation — Get up and running locally
- Environment Variables — Configure your environment
- Development Workflow — Daily development commands and tips
- Project Structure — Understand the codebase layout