The MosquitOff project was fully conceptualized, coded, and implemented by the participant. Artificial intelligence tools were used as supportive assistants in research, prediction, and documentation.

Scientific research assistance: ChatGPT (OpenAI) was used to identify and summarize epidemiological literature on Aedes aegypti dynamics and dengue-climate interactions, helping to inform the weighted algorithm used for risk estimation.
Technical guidance: ChatGPT also provided support in structuring the NASA data ingestion pipeline, including GitHub Actions workflows, Earthdata authentication, and daily automation of satellite data retrieval.
Predictive modeling: The project incorporates a LLaMA-based model that processes the previous seven days of aggregated temperature, precipitation, and soil moisture data, alongside scraped short-term weather forecasts, to produce a 7-day dengue-risk projection layer. This model augments, but does not replace, the project’s core epidemiological algorithm developed by Marcos Codas.
Editorial support: ChatGPT was used to refine and verify technical descriptions and dataset attributions for accuracy and clarity in the submission text.
Video Voice Creation: I used a model to create a text-to-speech as I could not find a quiet enough space to record the voice over for the 30-second video.

All data fusion, algorithmic weighting, code implementation, and visualization logic were developed independently by the participant. AI tools served as analytical and editorial aids, not as generators of original code or scientific conclusions.

2025 NASA Space Apps Challenge

MosquitOFF