Communications Engineer, Production Network Performance, Amazon Leo

We are looking for an engineer who builds data pipelines and writes code to understand a live satellite network. Amazon Leo is an initiative to launch a constellation of Low Earth Orbit satellites that will provide low-latency, high-speed broadband connectivity to unserved and underserved communities around the world. You will monitor Amazon Leo"s live satellite network against what our simulations predict, and use what you find to improve the performance our customers receive. You will help build and run an automated pipeline that compares real-time system and network data with simulation predictions, finds where performance falls short, and helps drive each problem to resolution with the many teams involved. A problem can sit anywhere in the system: satellites and their optical inter-satellite links, gateways, user terminals, the satellite telemetry, tracking and command (TT&C) link, and the software that connects them. The team also supports satellite operations: launch and early orbit, gateway and TT&C station assignments, and the satellite communication links that carry them. The team takes part in an on-call rotation for problems that affect customers or operations.

You will join a new, interdisciplinary team that covers software and data pipelines, radio frequency (RF) and satellite communication link performance (user, gateway and TT&C links), software and data pipelines, artificial intelligence and machine learning (AI/ML) automation, networking, systems integration and test, and production operations and troubleshooting, and that works closely with the modeling and simulation, hardware, software and operations teams across Amazon Leo.

Export Control Requirement

Due to applicable export control laws and regulations, candidates must be a U.S. citizen or national, U.S. permanent resident that is a current Green Card holder, or lawfully admitted into the U.S. as a refugee or granted asylum.

Key job responsibilities

Data pipeline and performance framework

• Build, maintain and use the team"s automated pipeline that brings live system and network data, simulation predictions and test measurements together
• Design how the data is stored, queried and kept at scale, so that the pipeline runs reliably and unattended across the whole network
• Write the code that joins sources logged at different rates and times, and keep each record"s context with it so that later analysis can explain it
• Integrate simulations and models developed by other teams into the team"s framework, and co-develop each tool"s automation interface with the team that owns it
• Compare live performance with simulation predictions within tolerances that reflect what each prediction means, and report when the model, not the network, looks wrong

Rapid detection of changes that affect customers

• Build AI, machine learning and automation steps into the pipeline to rapidly detect, group and triage changes that affect customers, and check their conclusions against independent evidence
• Develop statistical change-detection methods that flag when performance moves away from the network"s own history
• Turn large telemetry data sets into trends, visualizations and findings the team and its partners can act on

Troubleshooting and resolution

• Help trace performance problems to their root causes across RF, software, hardware and networking, by comparing affected and unaffected terminals, satellites and gateways
• Confirm a cause with an independent measurement, or reproduce the fault in simulation or test
• Hand each cause, with its evidence, to the team that owns the fix, and verify after the fix that customers see the improvement

Documentation and collaboration

• Build up a database of anomaly signatures, so that each investigation builds on earlier ones and known problems are recognized quickly
• Work with the simulation, test, satellite, gateway, terminal and software teams across our organization to get the data and evidence each investigation needs

A day in the life

About the team

We are a new team inside Amazon Leo's communication systems organization. We turn live network data into clear answers about the performance customers receive, and then work with subject matter experts to improve the customers' experience. Amazon Leo's network serves consumers, businesses and government organizations; within Amazon Leo, the team supports the engineering and operations teams that need reliable evidence of how the network performs.