EmberMind: A Generative AI chatbot solution for fire codes and safety information

The Problem

Fire prevention officers, supervisors, firefighting personnel, fire engineers and the public currently rely on cumbersome paper or electronic versions of complex fire codes, leading to significant challenges. This reliance creates inconsistencies in code interpretation, delays access to critical information during incidents, inspections and planning, and places a heavy cognitive burden on those who must retain and apply this knowledge. In the field, firefighters and fire prevention officers often struggle to access immediate, accurate code information, potentially compromising safety and efficiency. Furthermore, the current methods lack the ability to provide nuanced and context-aware responses to complex inquiries, which is a critical need that traditional information retrieval systems cannot adequately address.

The Solution

EmberMind is a Generative AI-powered tool designed to centralize and intelligently process a vast database of complex fire codes and safety information. By leveraging large language models like Gemini, the solution provides real-time, accurate and context-aware responses to complex inquiries through an interactive chatbot interface. EmberMind was built using open-source tools at no additional cost to the City.

The Impact

The tool centralizes complex fire codes and safety documents, transitioning staff away from manual, time-consuming lookups in cumbersome paper or electronic files.

Lessons Learned

Using NotebookLM as a prototype provided immediate insights into whether the model could provide high-quality, relevant answers from 2023 fire code standards without first investing in a custom user interface.

Launch year: 2025