Engineering Manager, AI Agents — Lead Production ML & DevEx

Requirements

• Led the technical delivery of complex, ambiguous products or systems in a domain without established playbooks, breaking down unknowns into testable hypotheses that led to meaningful shipped outcomes.
• Built and iterated on a structured experimentation or discovery framework (for example, spikes, A/B tests, or eval-driven loops) that reduced feasibility risk and accelerated decision making.
• Owned and delivered a DevEx or platform roadmap that materially improved engineering velocity, such as introducing CI/CD, improving testing and observability, or simplifying developer workflows.
• Managed or technically led a high-performing engineering team, creating psychological safety while holding a high bar for quality, delivery, and ownership of outcomes.
• Shipped AI / ML powered features or systems at production scale, working through challenges around reliability, observability, cost, and user trust, not just model performance in isolation.
• Collaborated closely with cross-functional partners (such as Product, Design, and Research / Data Science) to define strategy, shape roadmaps, and make trade-offs between delivery, debt, and foundational investment.

What the job involves

• Our Agent teams are building and scaling cutting-edge AI agents that power real customer workflows, from autonomous analysis to interactive assistance and user-facing automation.
• You will sit at the heart of translating ambitious product ideas and performance targets into robust, production-ready agent capabilities that materially move the needle for our customers.
• This group operates like a highly aligned, elite squad: tight collaboration between Engineering, Product, Design, and Research Engineering; fast feedback loops; and a bias towards experimentation over theory.
• Everything we build is novel, so we treat uncertainty as a design constraint, using formal explore / exploit practices to discover what works, kill what does not, and double down on impact.
• As Engineering Manager, you will own the technical delivery of one of our Agent roadmaps (Actuarial Agent, Ingestion Agent, or Underwriting Agent), leading AI Engineers to ship features that meet real customer requirements and close gaps surfaced through our eval pipelines.
• Your impact will be visible in the capabilities customers rely on day to day, and in the compounding velocity of a team that can repeatedly solve problems no one has solved before.
• Owning end-to-end technical delivery for one Agent team, turning customer requirements and eval-driven performance gaps into shipped capabilities that measurably improve agent success rates, latency, and reliability.
• Designing and institutionalising an explore / exploit operating model for your team, using timeboxed spikes, clear kill criteria, and exploitation triggers to reduce feasibility risk and increase the hit-rate of successful agent solutions.
• Partnering with Product, Design, and Research Engineering to co-design solutions, rapidly assessing feasibility for novel use cases, and steering the roadmap as you learn from real-world performance and customer feedback.
• Building and maintaining a DevEx roadmap focused on engineering velocity, investing in tooling, CI/CD, testing infrastructure, and observability so the team can safely increase deployment frequency and reduce cycle times.
• Coaching and mentoring AI Engineers to take on larger, more complex problems, growing technical leaders who can confidently drive designs, lead research spikes, and own critical areas of the agent stack.
• Creating a culture of high trust and high accountability, where experimentation is encouraged, data and evals guide decisions, and the team takes collective responsibility for the quality and performance of agents in production.