Overview:

In the first iteration, an LLM AI Data Analysis Engine was deployed to interact with a data room of climate action plans and supporting documents. The engine provides users with the ability to 'Ask' any question to one or all of the documents in the data room. The LLM AI then extracts a response from the document(s) and provides references showing where in the document the response came from.

The tool also features a 'Comparison' function, allowing users to select multiple Climate Action Plan (CAP) files and identify Cross-Over Zones. These zones represent opportunities for synergies and potential conflicts between different climate action plans. The tool responds with all the identified cross-over zones, explains why they are cross-over zones, and suggests potential actions that the council in charge can take to address them. It also highlights the councils involved.
‍

Key Components:

• Developed an LLM-powered 'Ask' feature for intuitive interaction with the documents
• Created an LLM-driven 'Comparison' feature to identify 'Cross-Over Zones' and related clauses between many 40+ page long documents.
• The system is designed to continuously evolve and adapt based on user interactions and feedback
• Fine-tuned the LLM AI for improved accuracy in extracting responses and identifying cross-over zones

‍