Igor May Wensing

Background

About Igor May Wensing

I am currently a Data Scientist with a focus on LLMops at the Court of Auditors of the State of Santa Catarina (TCE-SC). I play a key role in the development of the VigIA project, which utilizes LLMs and Retrieval Augmented Generation (RAG) alongside a robust Big Data infrastructure to automate the analysis of public documents, identify irregularities, and optimize public spending.One of my main responsibilities is optimizing our application that uses LLM and RAG at large-scale processing. I work directly on the implementation and optimization of RAG, focusing on improving performance, reducing costs, decreasing response times, and increasing the application's efficiency. In my daily work, I use a variety of technologies, including Apache Spark, Airflow, Kafka, Hive, Impala, SQL, Python, LLMs, LLM Fine-tuning, Prompt Engineering, GenAI code assistants (such as Github Copilot), Software Architecture, Machine Learning, AI Solution Prototyping, TensorFlow, PyTorch, Scikit-learn, Pandas, NumPy, ETL (Extract, Transform, Load), Data Wrangling, Data Visualization (using tools like Matplotlib, Seaborn, Tableau), Docker, Kubernetes, and CI/CD, among others.Previously, I worked as a Software Engineer at GE and hold a degree in Computer Science from the Federal University of Santa Catarina. My experience encompasses implementing innovative solutions using generative AI and LLMs to solve complex business problems, always focusing on increasing efficiency, reducing human errors, and improving code quality.