Joao Vitor Valera

At Nava, I work as a Senior MLOps Engineer supporting data teams in Azure environments, with a focus on experiment traceability, standardized model lifecycles, and data-driven culture.

Previously at Accenture Brasil, I developed and deployed machine learning solutions for retail, telecommunications, and mining clients, including demand forecasting, churn, and resource-allocation models. I built probabilistic B2B order forecasts to optimize delivery allocation and reduce operational losses, while creating automated time-series pipelines with Airflow and MLflow.

I combine hands-on data science with MLOps practices across Python, SQL, MLflow, Docker, Azure, AWS, Databricks, and PySpark to make ML reliable, auditable, and sustainable in production.