Summary
Lorenzo Cesconetto is a versatile software engineer with nine years of experience building scalable, event-driven systems and ML-driven products across cloud-native and research environments. He blends hands-on expertise in Python, transformers, and distributed training with production-grade serverless and microservices architectures on AWS and GCP, having driven major resiliency and cost improvements at Cellanome and Deepcell. Lorenzo has shipped NLP models for Portuguese (NeoBERTugues) and end-to-end RAG/semantic search solutions for enterprise chatbots, plus an accessible Python SDK and data pipelines that cut data loss to nearly zero. He pairs a strong mathematical background and academic training from ITA with practical startup chops—founding and exiting a ChatGPT customization product—and publishes technical deep dives on Transformers, LLMs, and GPUs. Based in Brazil, he’s equally comfortable optimizing Postgres queries as orchestrating large-scale PyTorch distributed runs, and is known for pragmatic innovations like smart weight initialization and hybrid dense-sparse retrieval strategies.
9 years of coding experience
6 years of employment as a software developer
Pós Graduação, Ciência de Dados - Data Science, Pós Graduação, Ciência de Dados - Data Science at Instituto Tecnológico de Aeronáutica - ITA
Hero!, Entrepreneurship/Entrepreneurial Studies, Hero!, Entrepreneurship/Entrepreneurial Studies at Draper University
Exchange Student, Civil Engineering, Exchange Student, Civil Engineering at University of Nebraska-Lincoln
Civil Engineering, Engenharia Civil, Civil Engineering, Engenharia Civil at Universidade Estadual de Campinas
Civil Engineering, Civil Engineering at Universidade Federal do Espírito Santo
Expert Solidity Bootcamp, Computer Science, Expert Solidity Bootcamp, Computer Science at Encode Club
Entrepreneurship/Entrepreneurial Studies, Entrepreneurship/Entrepreneurial Studies at Latitud
English, Portuguese