Jobs in Sao Paulo, United States

We’re seeking an experienced Machine Learning Engineering Manager to lead cross‑functional teams in designing, training, and deploying large‑scale ML and LLM systems. As a hands‑on leader, you'll drive the full lifecycle of AI development— from research and experimentation to distributed training and production deployment—while mentoring top‑tier engineers and partnering closely with product, research, and infra leaders.

This role blends deep ML/MLOps expertise with strong leadership and execution, ensuring all AI initiatives translate into measurable business impact.

• Lead and mentor ML engineers, data scientists, and MLOps professionals.


• Manage end‑to‑end ML/LLM project lifecycle: data pipelines, training, evaluation, deployment, and monitoring.


• Provide technical direction for distributed training, large‑scale model optimization, and system architecture.


• Collaborate with Research, Product, and Infrastructure teams to define objectives, milestones, and KPIs.


• Implement MLOps best practices: experiment tracking, CI/CD, model governance, observability.


• Manage compute resources, enforce Responsible AI + data security standards.


• Communicate technical progress, blockers, and results clearly to leadership and stakeholders.

• 5+ years of experience in Machine Learning, NLP, and Deep Learning (Transformers, LLMs).


• 2+ years leading teams delivering ML/LLM systems in production.


• Strong proficiency in Python and frameworks like PyTorch, TensorFlow, Hugging Face, DeepSpeed.


• Experience with distributed training, GPU/TPU optimization, and cloud platforms (AWS, GCP, Azure).


• Knowledge of MLOps tools (MLflow, Kubeflow, Vertex AI, etc.).


• Excellent leadership, communication, and cross‑functional collaboration skills.


• Bachelor’s / Master’s in Computer Science, Engineering, or related field (PhD preferred).

• Experience training or fine‑tuning foundation models.


• Contributions to open‑source ML/LLM frameworks.


• Knowledge of Responsible AI practices, bias mitigation, and model interpretability.