Senior Engineering Manager, Machine Learning

Description

The Custom Agents group at the company lets customers build specialized AI agents. These agents understand business context, reason through complex tasks, and take actions across the company's product and connected systems. They handle work such as investigating requests, reading documents, applying policies, and coordinating service workflows.

The group's aim is to make these agents reliable, configurable, and useful in real business operations.

This is a senior management role leading the machine learning work behind Custom Agents. You will set the technical direction, build a strong team, and turn progress in generative AI into dependable customer features.

The role combines people leadership, applied ML judgment, and responsibility for shipping to production. You will work closely with Product, Engineering, Design, ML Science, and the Agent Platform team to improve how agents reason, use tools, follow policies, and finish tasks. You will weigh model quality, reliability, latency, and cost, and tie technical work to measurable customer value.

Responsibilities

• Hire, coach, and develop ML engineers and scientists, and grow future leaders within the team. Foster ownership, careful experimentation, inclusion, and high technical standards.
• Set the ML strategy and roadmap. Turn customer needs and product goals into clear priorities, investment choices, and realistic milestones. Balance foundational work with shipping new capabilities.
• Improve how agents plan, choose tools, manage context, ground answers in knowledge, and produce structured outputs so they reliably complete the intended task.
• Build evaluation practices: representative benchmarks, human annotation, automated scorers, and regression tests. Measure task completion, correct tool use, policy adherence, and whether reported results match what the agent actually did.
• Set standards for experiments, validation, release readiness, and production monitoring. Use execution traces and customer feedback to find causes of failure and drive fixes.
• Compare models and agent designs on quality, latency, and cost. Decide on model selection, routing, fallbacks, and inference efficiency.
• Work with Security and Privacy teams on permission-aware actions, defenses against prompt injection, data handling, and escalation or human approval.
• Adopt and help shape shared platform tools for simulation, observability, experimentation, and durable execution. Work with adjacent teams to fit Custom Agents into wider service workflows.
• Link ML improvements to automation success, shorter handling time, customer adoption, and trust. Report progress, tradeoffs, and risks clearly to stakeholders.

Requirements

• A proven record of leading ML or AI teams and delivering production systems with measurable customer impact.
• Strong people management, including hiring, coaching, performance management, and growing technical leaders.
• Solid understanding of modern ML and generative AI, with hands-on experience moving LLM applications from experiments into production.
• Good technical judgment across model behavior, software architecture, evaluation, and operational reliability. You can lead design reviews and assess experimental results.
• Experience defining useful quality metrics and using data to set priorities and make release decisions.
• Ability to lead through ambiguity, coordinate several workstreams, and influence partners in Product, Engineering, Science, and Security.
• Clear communication and a practical approach that balances scientific rigor, speed of delivery, and customer needs.
• Additional experience we value: building agents that use tools, retrieve enterprise knowledge, or coordinate workflows across systems.
• Additional experience we value: working with LLM evaluation, human feedback, multimodal document processing, or agent safety.
• Additional experience we value: leading distributed teams or delivering ML products in enterprise SaaS, customer service, or business automation.

Conditions

Hybrid role: you will spend part of the week in person at one of the company's offices and may work remotely for the rest.

Applicants are told that artificial intelligence or automated decision systems may be used to screen or evaluate applications, in line with company guidelines and applicable law.