Summary
Vinicius Arruda is a Natural Language Processing Engineer with eight years of hands-on experience building ML systems that bridge research and production. He has designed LLM agents, RAG-based retrieval pipelines, and rule-based NLP tools to extract and generate engineering requirements, working with models from GPT-3.5/GPT-4 to Falcon, Mistral and LLaMA variants. His background spans full-stack development, DevOps and embedded TTS pipelines for pt-BR, giving him a rare combination of applied ML research, speech engineering and production deployment skills. He holds an MS in Computer Science with top grades and contributed to academic work on domain adaptation and synthetic data generation during his fellowship. Comfortable across Python, PyTorch, TensorFlow and Hugging Face tooling, he also brings experience integrating ML into devices (Jetson) and cloud services. Colleagues describe him as a pragmatic researcher who translates complex language and vision models into usable product features.
8 years of coding experience
2 years of employment as a software developer
Bachelor's degree, Civil Engineering, Bachelor's degree, Civil Engineering at Universidade Vila Velha - UVV
Técnico em Informática, Técnico em Informática at Ifes - Instituto Federal do Espírito Santo
Master of Science - MS, Computer Science, 9.38/10.0, Master of Science - MS, Computer Science, 9.38/10.0 at Ufes - Universidade Federal do Espírito Santo
Portuguese, English