Sr. Manager, Software Development, Prime Video

Prime Video operates one of the largest and most complex streaming infrastructures in the world, serving hundreds of millions of customers across live sports, movies, and TV in more than 240 countries and territories. Ensuring that this infrastructure scales predictably, runs efficiently, and protects customer privacy is foundational to the business.

This leader will report directly to the Director of READI and be accountable for how Prime Video plans for scale, drives cost and resource efficiency across core compute and AI infrastructure, and upholds customer privacy as a first-class engineering discipline. This is a broad role that spans predictive capacity planning, deep infrastructure optimization, and privacy engineering — requiring a leader who is equally comfortable in a data-driven forecasting review, an architecture deep-dive on compute and GPU efficiency, and a privacy risk assessment with legal and security partners.

The successful candidate is a seasoned engineering leader who builds and develops high-performing teams, sets multi-year technical strategy, and delivers measurable results in ambiguous, high-scale environments. They will lead through other managers and senior engineers, influence peer organizations and partner teams across Prime Video and AWS, and represent Prime Video’s infrastructure posture to senior leadership.

• Own the end-to-end discipline of forecasting demand and assuring capacity across Prime Video’s infrastructure footprint, ensuring the service scales reliably for both predictable growth and peak events (e.g., major live sports, tentpole launches, and global releases).
• Own multi-year and event-based capacity models that translate business growth, content strategy, and traffic patterns into infrastructure demand across compute, storage, network, and CDN.
• Lead capacity assurance for high-stakes live events, establishing headroom targets, load-testing regimes, and go/no-go readiness reviews with partner teams.
• Build a forecasting practice grounded in data science and telemetry — improving forecast accuracy, quantifying confidence intervals, and reducing both over-provisioning waste and under-provisioning risk.
• Partner on demand-planning and capacity-management processes with AWS, finance, and PV service teams so that capacity is secured ahead of need at the right cost.
• Define the mechanisms (dashboards, reviews, and escalation paths) that give leadership confidence that Prime Video will scale for every moment that matters.

• Drive infrastructure efficiency as a durable, measurable engineering discipline across Prime Video’s core compute fleet and its rapidly growing AI/ML infrastructure, materially improving unit economics while protecting reliability and performance.
• Set the efficiency strategy and roadmap for core compute (CPU fleets, containerized services, memory/DRAM footprint) and AI infrastructure (GPU/accelerator fleets, training and inference platforms).
• Drive down infrastructure cost-to-serve and improve utilization through right-sizing, fleet consolidation, hardware refresh strategy, and adoption of more efficient compute and runtimes.
• Optimize the efficiency of AI/ML workloads — GPU utilization, training throughput, inference cost per request — partnering with science and platform teams to eliminate waste in one of PV’s fastest-growing spend areas.
• Establish efficiency goals with clear metrics (utilization, cost per stream/request, DRAM and energy footprint) and instrument the fleet so gains are visible, attributable, and durable.
• Lead zero-based capacity and efficiency reviews, and represent PV in senior compute-allocation and…