ML/AI Production Systems Manager

Socket.dev; Bellevue (WA); On-site; USD 185,000 - 250,000; Full time

Benefits offered by this job

Health insurance

401(k) matching

Parental leave

Paid time off

Job summary

Amazon's PXTCS team seeks a Software Development Manager to lead an engineering group responsible for taking ML/AI models from research to reliable production systems. You will own the technical strategy, architecture, and operational excellence of services hosting these models while growing the team and mentoring engineers.

You will collaborate with applied scientists, economists, and data scientists to translate prototypes into scalable, secure, and fault‑tolerant systems, aligning with TPMs

Qualifications

• 7+ years of engineering experience and 3+ years of engineering team management.
• Experience leading all aspects of the software development lifecycle.
• Hands-on with large-scale distributed systems and multi-tier architectures.

Responsibilities

• Own the technical strategy, architecture, and end-to-end delivery of production services for ML/AI models.
• Hire, mentor, and grow a high-performing software engineering team.
• Collaborate with scientists, data engineers, and product teams to productionize research.
• Set engineering standards and drive operational excellence including monitoring and incident response.
• Communicate tradeoffs and status to leadership and stakeholders.

Skills

Engineering leadership

Distributed systems

Software lifecycle

Team management

Stakeholder communication

Hiring & mentoring

Education

Bachelor's degree in CS/Engineering/Math

Job description

Amazon's PXTCS team seeks a Software Development Manager to lead an engineering group responsible for taking ML/AI models from research to reliable production systems. You will own the technical strategy, architecture, and operational excellence of services hosting these models while growing the team and mentoring engineers.

You will collaborate with applied scientists, economists, and data scientists to translate prototypes into scalable, secure, and fault‑tolerant systems, aligning with TPMs