Summary
Walter Augusto is an AI-focused data scientist and postgraduate researcher with 8 years of experience building production-ready ML and MLOps solutions across cloud providers (AWS, GCP, OCI, Azure). He designs and implements multi-agent LLM architectures and fast inference engines using LangChain, DeepSeek, LLaMA-family models, TensorFlow and PyTorch, and integrates LLM-driven bots into platforms like Telegram to automate workflows and intelligent interactions. His work spans finance-focused intelligent query systems, vector DBs (ChromaDB, FAISS) for semantic search, and bioinformatics experiments such as molecule generation with VAE+SMILES, reflecting a rare mix of applied NLP, computational neuroscience and engineering neuromorphic concepts. Comfortable from low-level architectures to experiment tracking (MLflow) and cloud orchestration, he combines strong operational roots (SysOps/SRE, DevOps) with academic rigor at ICMC-USP, bringing reproducibility and scalability to complex AI research-to-production pipelines.
8 years of coding experience
4 years of employment as a software developer
Bacharelado incompleto, Física, Bacharelado incompleto, Física at USP - Universidade de São Paulo
Bacharel em Ciência de Dados, Ciência de dados, Bacharel em Ciência de Dados, Ciência de dados at Univesp Oficial
Spanish, English, Russian