ProcureNow is a high-end, AI-powered platform designed to automate the compliance auditing of construction proposals against Request for Proposal (RFP) requirements. Using a 4-phase multimodal pipeline driven by Google Gemini AI, it identifies compliance gaps, evaluates risk factors, and provides a structured audit report with high precision.
- Multimodal PDF Ingestion: Processes complex documents (tables, charts, grids) using Gemini Vision to ensure no data is missed.
- Dynamic RFP Rubric Extraction: Automatically identifies and extracts requirements from the RFP to build a custom compliance rubric.
- Section-Routed Audit: Semantically routes requirements to the relevant sections of the proposal for accurate presence and accuracy evaluation.
- Interactive Dashboard: A professional, data-dense interface for visualizing overall compliance, critical focus areas, and detailed requirement status.
- Evidence Viewer: A side-by-side viewer that allows users to verify AI-extracted findings directly against the source proposal PDF.
- Branded Exports: Generate professional PDF audit reports or export results to CSV for further analysis.
- Backend: FastAPI, Uvicorn, Python 3.11+
- AI/LLM: Google Gemini Pro (Multimodal)
- PDF Engine: PyMuPDF (fitz), fpdf2
- Frontend: Vanilla JavaScript (ES6+), Modern CSS (Flexible Layout, Glassmorphism), Semantic HTML5
- Database: SQLite (via standard Python library)
ProcureNow/
├── backend/ # FastAPI application logic
│ ├── auditor.py # Core compliance auditing logic
│ ├── extractor.py # RFP requirement extraction
│ ├── server.py # Entry point and API routes
│ └── ... # Helpers (PDF reading, database, reporting)
├── frontend/ # Static dashboard files
│ ├── index.html # Main dashboard structure
│ ├── app.js # Frontend logic and API integration
│ └── index.css # Styling (Forest & Lime theme)
├── data/ # Local database storage
├── design-system.md # Detailed Design Tokens and UI/UX rules
├── requirements.txt # Python dependencies
└── .env # Environment variables (API Keys)
- Python 3.11 or higher installed on your system.
- A Google Gemini API Key (available via Google AI Studio).
pip install -r requirements.txtCreate a .env file in the root directory (or update the existing one):
GEMINI_API_KEY=your_gemini_api_key_hereStart the server using the following command:
python3 -m backend.serverOnce the server is running, access the dashboard at: http://localhost:8000
The project follows a strict brand guideline defined in design-system.md. It utilizes a professional Forest Green and Lime palette, prioritizing readability, data density, and a "No-Line" visual philosophy.
