Nexa’s core challenge is scaling model development and production operations together so that new AI capabilities remain fast, reliable, and maintainable as the product grows. At Alighieri, I lead production AI delivery across PyTorch training, data pipelines, FastAPI services, Kubernetes deployments, monitoring, and optimization. This experience would help Nexa expand its AI capabilities while keeping model development closely integrated with the software and infrastructure required to operate them reliably. My current work includes configurable PyTorch training pipelines with dataset validation, experiment tracking, checkpoint management, hyperparameter tuning, and reproducible evaluation. I build Python, SQL, Spark, and Airflow pipelines to process structured and unstructured data from databases, APIs, and object storage. MLflow is used to track experiments and model versions and to support controlled promotion between development and production environments. For model serving, I build FastAPI services for real-time and asynchronous inference, incorporating validation, batching, timeouts, retries, health checks, and clear failure handling. Docker and Kubernetes support deployments across AWS and GCP. I optimize runtime performance through mixed precision, quantization, model warm-up, caching, request batching, resource profiling, and autoscaling. Monitoring covers model quality, data drift, latency, service errors, and infrastructure health. As a technical lead, I make architecture decisions, review code, mentor engineers, and collaborate with product, software, and data teams to integrate models into complete applications. I'm comfortable working with US-based and international stakeholders and explaining technical tradeoffs in clear, practical terms. I would welcome the opportunity to discuss how my experience can support Nexa's production AI roadmap.…Nexa’s core challenge is scaling model development and production operations together so that new AI capabilities remain fast, reliable, and maintainable as the product grows. At Alighieri, I lead production AI delivery across PyTorch training, data pipelines, FastAPI services, Kubernetes deployments, monitoring, and optimization. This experience would help Nexa expand its AI capabilities while keeping model development closely integrated with the software and infrastructure required to operate them reliably. My current work includes configurable PyTorch training pipelines with dataset validation, experiment tracking, checkpoint management, hyperparameter tuning, and reproducible evaluation. I build Python, SQL, Spark, and Airflow pipelines to process structured and unstructured data from databases, APIs, and object storage. MLflow is used to track experiments and model versions and to support controlled promotion between development and production environments. For model serving, I build FastAPI services for real-time and asynchronous inference, incorporating validation, batching, timeouts, retries, health checks, and clear failure handling. Docker and Kubernetes support deployments across AWS and GCP. I optimize runtime performance through mixed precision, quantization, model warm-up, caching, request batching, resource profiling, and autoscaling. Monitoring covers model quality, data drift, latency, service errors, and infrastructure health. As a technical lead, I make architecture decisions, review code, mentor engineers, and collaborate with product, software, and data teams to integrate models into complete applications. I'm comfortable working with US-based and international stakeholders and explaining technical tradeoffs in clear, practical terms. I would welcome the opportunity to discuss how my experience can support Nexa's production AI roadmap.WWWW…