AI-Powered DDR Generation System (Mistral AI + claude +different models)

📌 Project Objective

This system automates the conversion of raw site inspection and thermal data into a structured, client-ready Detailed Diagnostic Report (DDR)

🛠️ Tech Stack

• LLM: Mistral AI (Reasoning, Deduplication, and Synthesis) + claude
• Orchestration: LangChain
• Extraction: PyMuPDF / Unstructured (Text & Image extraction)
• Language: Python 3.10+

🚀 Key Features

• Multimodal Data Merging: Combines general site observations with thermal imaging data.
• Conflict & Gap Handling: Explicitly identifies missing information as "Not Available" and flags data conflicts
• Severity Assessment: Uses AI reasoning to assign severity levels with clear logic.
• Image Integration: Automatically places relevant inspection images into corresponding report sections.

⚙️ Setup & Installation

1. Clone the repository: git clone https://github.com/Ansh7473/Ai-Generalist
2. Install dependencies: pip install -r requirements.txt
3. Configure Environment: Create a .env file based on .env.example and add your MISTRAL_API_KEY.
4. Run the generator: python src/main.py