Senior Applied Scientist, Network Fabric Engineering

We are looking for an Applied Scientist to join our team and tackle some of the most challenging traffic engineering problems at planetary scale. You will develop novel optimization algorithms

and ML models that balance network utilization, resilience, cost efficiency, and latency — across traffic classes with fundamentally different characteristics. The problem then expands to network capacity planning, where you will collaborate with other scientists to develop the optimal strategy to efficiently scale the network.

Key job responsibilities

• Formulate traffic engineering problems as mathematical optimization models and develop scalable solvers that operate on graphs with hundreds of nodes and O(10^4) edges
• Design ML models for traffic demand prediction, anomaly detection, and workload classification
• Develop capacity planning frameworks that co-optimize cost, reliability, and performance over multi-year horizons
• Build simulation and evaluation frameworks to validate TE algorithms against realistic failure scenarios and traffic patterns
• Publish research at top venues (SIGCOMM, NSDI, INFOCOM, NeurIPS, ICML) and contribute to the scientific community
• Collaborate with network engineers, software engineers, and operations teams to bring algorithms from prototype to production at global scale
• Define metrics, run A/B experiments, and measure the real-world impact of algorithmic changes on network performance

About the team

The Inter Data Center (InterDC) Networking team owns traffic engineering across AWS's regions — the largest inter-datacenter network in the world, connecting hundreds of data centers

across dozens of regions. We design and implement the algorithms, optimization systems, and scientific frameworks that decide how exabytes of traffic traverse this network every day. Our work

sits at the intersection of combinatorial optimization, machine learning, distributed systems, and network engineering.