rafaeltuelho/construction-spec-assistant
🤖 AI Agentrafaeltuelho
An agentic AI system for architects to automatically detect inconsistencies between construction specifications and contractor submittals.
The Construction Spec Assistant is a sophisticated agentic workflow built using Python and Jupyter Notebooks, specifically engineered for the Architecture, Engineering, and Construction (AEC) sector. At its core, the system utilizes Docling for robust document parsing and OCR, enabling it to ingest complex construction specifications and technical drawings. By integrating LangGraph and LangChain, the assistant orchestrates a multi-step RAG (Retrieval-Augmented Generation) pipeline that cross-references contractor submittals against the master project requirements. This automation significantly reduces the manual labor involved in quality assurance and compliance checking. The system is designed to highlight inconsistencies, missing information, or deviations from original design intent, providing architects with actionable insights to mitigate project risks and prevent costly construction errors. Its modular architecture allows for future expansion into various document types and regulatory standards.
💡Highlights
- ├─Agentic RAG for document comparison
- ├─Docling-powered OCR and parsing
- └─Automated discrepancy detection
🎯For
- ├─Construction Architects
- ├─Civil Engineers
- └─Project Managers