Students Advance to the Finals of an Energy Transition Hackathon

How can artificial intelligence help address the challenges of the energy transition? It was in search of answers to this question that five students from IMPA Tech participated in the Coppe 2026 AI Hackathon, organized by the Coppe/UFRJ Business Incubator and Made in Rio. The group advanced to the in-person stage of the competition, held September 25–27, and ranked among the top 12 teams, competing against an audience of undergraduate and graduate students, researchers, and professionals from various fields.

Composed of Igor Roberto Alves, Letícia Aleixo, Pedro Henrique Porto, Pedro Alberti, and Richard Elias—all from the class of 2024 and specializing in Artificial Intelligence and Data Science— the group worked on a challenge proposed by Taesa (Transmissora Aliança de Energia Elétrica S.A.): to develop a solution for competitive intelligence in the electric power sector, with a focus on data processing and usability.

The challenge stemmed from a common issue in the industry: energy companies disclose information on sustainability, performance, and financial results, but this data may be scattered across different documents and formats. The students’ proposal was to use artificial intelligence to organize this content and make it more accessible.

The competition began with the selection of 120 participants based on an analysis of their résumés. From September 11 to 13, the participants took part in the Ideathon, a phase dedicated to developing proposals for the challenges presented. At the end of this phase, 12 groups advanced to the next stage.

It was at that moment that the IMPA Tech students began developing the solution for Taesa. The group worked on the project and, during the in-person phase, had the opportunity to present and validate the tool directly with professionals in the field.

“In the energy sector, all companies publish their sustainability and financial performance data, but the problem is that they can report this in any format—not necessarily on an integrated platform,” explains Igor. “What we did was bring together the largest companies, collect the available data, and derive insights from that material, transforming the PDFs into accessible information.”

Based on the data collected, the group developed a functional platform with a responsive chatbot capable of facilitating the retrieval and analysis of information. Despite their mathematical skills, the challenge lay in understanding content specific to the energy sector. “We couldn’t grasp all the terminology in the field, which made it difficult to evaluate the decisions. We lacked a feel for the market,” he explained.

The experience also led the students to reflect on the role of the data scientist in projects aimed at solving real-world problems. “These days, a data scientist can’t just sit behind a computer. You have to understand the client, the field of work, understand the nature of the data, and know how to sell the product,” he emphasizes. “The more we expand our horizons and our professional network, the greater the possibilities for applying what we’ve learned and making a real social impact,” he added.