Staff Engineer, Machine Learning

About the Role

Over 5 billion users worldwide use basic applications like email, notes, tasks, and calendars without native AI integration. Our goal is to create proactive, intelligent applications requiring minimal or no complex prompts, enhancing conversations, errands, organization, and workflows with greater reliability and real-world task completion while reducing hallucinations.

We want to help organize users' lives, freeing up time for more valuable and meaningful activities. As a Staff Engineer in Machine Learning, you will lead the execution of ML intelligence, translating research models into dependable, scalable production systems.

Key Responsibilities

• Manage end-to-end ML systems covering data handling, training, evaluation, inference, and deployment.
• Develop and improve pipelines for training and fine-tuning large-scale models.
• Create evaluation frameworks assessing model capabilities, robustness, safety, and real-world effectiveness.
• Design and optimize high-performance inference systems balancing latency, GPU use, memory, cost, and stability.
• Develop data pipelines for high-quality real-world and synthetic training datasets.
• Build dependable production infrastructure for model deployment, monitoring, and continual enhancement.
• Collaborate closely with research and application engineering to integrate model competencies into product upgrades.
• Make informed technical trade-offs and rapidly iterate based on performance data.

Required Qualifications

• Proven experience deploying ML systems in production environments, beyond experimental prototypes.
• In-depth understanding of contemporary large model training, fine-tuning, evaluation, and inference techniques.
• Strong software engineering and system fundamentals.
• Experience managing scalable ML workloads, especially GPU-oriented systems.
• Excellent technical judgment and the ability to handle ambiguous challenges independently.
• A strong preference for experimentation, data-driven decision-making, and timely delivery.
• Commitment to high standards in accuracy, reliability, and production-level quality.

Expected Outcomes

• Seamless translation of research and models into production-ready solutions with measurable quality benchmarks.
• Stable, efficient, and maintainable ML pipelines, training processes, and inference infrastructures.
• Rapid identification and resolution of production issues with minimized user disruption.
• Supportive, aligned teams capable of delivering impactful ML solutions with minimal obstacles.
• Continuous, measurable improvements in models and systems enhancing user experience safely over time.

Technical Environment

• Programming in Python
• Use of ML frameworks such as PyTorch and JAX
• GPU-based training and inference platforms

Preferred Experience

• Building or shipping real-world ML systems affecting users, not just prototypes.
• Comfort with large model complexities and understanding their failure mechanisms.
• Strong coding skills with a focus on production-grade system reliability.

Work Culture

Our team is compact but highly skilled, encouraging engineers to own broad responsibilities and exercise strong independent judgment. We prioritize rapid decision-making and close collaboration, focusing on engineering excellence over rigid processes.

Interview Process

Candidates who fit the role will be invited for a maximum of four interviews, conducted virtually or onsite, evaluated by our technical team. We emphasize prompt and transparent communication, aiming to extend offers swiftly to exceptional candidates who share our vision of bringing practical AI benefits to billions globally.