Multimodal Chatbot with LLMs for Incident Handling in Public Safety Services

• Luis Henrique Miranda Queiroz UFPI
• Raimundo Santos Moura UFPI

DOI: https://doi.org/10.5753/enucompi.2025.9617

Abstract

This paper presents the development of an intelligent conversational agent, based on large-scale language models (LLMs), to assist in the registration of police reports and emergency services through digital platforms such as WhatsApp. The prototype allows multimodal communication (text, audio, image and location), automatically classifying occurrences, identifying gaps in reports and performing validations via external services. The solution uses technologies such as FastAPI, LangChain, Google Cloud Platform and Twilio, and is integrated with transcription, computer vision and data verification tools. The system is already functional in a controlled environment, with initial tests showing positive results. This work is part of the intersection between Artificial Intelligence, Human-Computer Interaction and Public Informatics, promoting accessibility, agility and security in communication between citizens and public security services.

References

Dam, S. K., Hong, C. S., Qiao, Y., and Zhang, C. (2024). A complete survey on llm-based ai chatbots. arXiv preprint.

Gunnam, G. R., Inupakutika, D., Mundlamuri, R., Kaghyan, S., and Akopian, D. (2024). Assessing performance of cloud-based heterogeneous chatbot systems and a case study. IEEE Access.

Oliveira, E. G. (2024). Chatbots: a importância do processamento de língua natural para a experiência do usuário. Master’s thesis, Universidade Federal do Rio Grande do Norte (UFRN).