An archaeological journey to unearth how LLMs work
Israel Thalles Dutra dos Santos presented at FCTE/UnB the TCC1 that designs a gamified web platform — framed as an archaeological exploration — to teach the foundations of large language models to Software Engineering students.
Brasília, July 2026
Large language models (LLMs) have become invisible infrastructure for both work and study in Software Engineering — but explaining how they actually work is rarely straightforward. It is this double gap — motivation and structured teaching material — that Israel Thalles Dutra dos Santos chose to tackle in his TCC1, presented at the Faculty of Sciences and Technologies in Engineering (FCTE), advised by Prof. Sergio Antônio Andrade de Freitas .
The proposal is a web platform aimed at teaching the fundamentals of LLMs — artificial intelligence, machine learning, deep learning, and natural language processing — designed as an archaeological exploration journey. The metaphor is not decoration: each stage of the game corresponds to a content block the student excavates, with curiosity, unpredictability, and a sense of epic mission as narrative drivers.
The methodological backbone is the Octalysis framework. Israel administered the Octalysis questionnaire to a group of target students, analyzed the data with the Cronbach alpha coefficient, and identified four dominant core drives: development and accomplishment, epic meaning, empowerment of creativity and feedback, and unpredictability and curiosity. From those findings came ten Gamification techniques selected for the platform — not chosen by taste, but grounded in the motivational profile actually observed.
The work brings together two fronts that often travel apart at CEDIS: artificial intelligence as the content to be taught and gamification as the instrument to teach it. TCC1 closes the foundations, requirements, and prototype phase — and sets the stage for TCC2, when the platform will be implemented and evaluated with real students.
About CEDIS: The Center for Studies, Development, and Innovation in Software (CEDIS), linked to the University of Brasília, researches and develops innovative software solutions with strong presence in Gamification and artificial intelligence .