João Gabriel Rossi de Borba

Conduct research on applied AI systems for translating ethical requirements and evaluating LLM behavior while supporting software engineering workflows. Build LLM- and RAG-based pipelines with structured prompting, automated evaluation, guardrails, fuzzing, runtime validation, and experiment logging.

Coordinated requirements, defined scope, and organized backlogs for internal tools supporting institutional processes. Led back-end delivery and implemented process automation solutions using Python, JavaScript, APIs, SQL databases, and cloud-based infrastructure.

Teach Computer Architecture and Algorithms, including CPU design, pipelining, I/O, parallelism, and core algorithm concepts. Deliver workshops and design instructional materials and learning activities to support student engagement.

Teach Discrete Mathematics (logic, sets, relations, and graphs) and Software Development Methods and Software Product Engineering with an emphasis on agile methodologies. Supervise capstone projects and support master’s-level supervision through instructional material design and academic contributions.

Designed and implemented AI solutions for judicial automation, leading a legal document processing project using NLP. Optimized models and processing pipelines, reducing execution time by 40%, and delivered evaluation routines for model performance.

Fixed bugs and improved production software while contributing to architectural improvements and new system features. Worked on systems used by telecom operators including Vivo, TIM, and Claro.

Led a team to develop a secure and scalable digital archive control system. Coordinated agile workflows and designed and implemented database solutions for large-scale information management.

Developed and maintained back-end features using Django. Refactored for improved performance and maintainability and optimized databases while building APIs.

Pursuing a Ph.D. in Computer Science researching Ethical Artificial Intelligence using domain-specific languages, agentic systems, and fuzzing. Focuses on operationalizing, validating, and governing ethical requirements in LLM-based systems, including guardrails/policy execution mechanisms and automated evaluation and runtime validation.

Completed an M.Sc. in Computer Science focusing on AI, LLMs, AI ethics, agentic systems, and software engineering. Developed an LLM-based tool translating ethical requirements into user stories using RAG, structured prompting, and evaluation, and explored MLOps/LLMOps for integrating generative models with systematic quality, cost, and behavior evaluation.

Completed a B.Sc. in Software Engineering with research on AI applied to software systems. Developed a web-based Quality of Work Life system for the Brazilian public sector.