Software Development Engineer, Applied AI Solutions

As part of the AWS Applied AI Solutions organization, we have a vision to provide business applications, leveraging Amazon’s unique experience and expertise, that are used by millions of companies worldwide to manage day-to-day operations. We will accomplish this by accelerating our customers’ businesses through delivery of intuitive and differentiated technology solutions that solve enduring business challenges. We blend vision with curiosity and Amazon’s real-world experience to build opinionated, turnkey solutions. Where customers prefer to buy over build, we become their trusted partner with solutions that are no-brainers to buy and easy to use.

AWS Physical AI is building the platform that helps companies developing autonomous systems validate their safety-critical applications. Our team builds solutions that automate the creation of test scenarios, generate synthetic sensor data, and enable simulation-based validation — covering edge cases and boundary conditions that are difficult to capture through real-world data collection alone.

Physical AI spans the full lifecycle of autonomous systems development — from data curation and model training through simulation-based validation and deployment monitoring. As a Software Development Engineer on the Physical AI team, you'll work across this stack, building core capabilities that transform operational design specifications into realistic synthetic scenarios, sensor data, and validation workflows.

You'll work with generative AI models trained on real-world operational data to create realistic agent behaviors, object interactions, and environmental conditions that systematically explore safety-critical situations. Your work will directly enable validation engineers at companies building autonomous systems to achieve comprehensive test coverage that previously required months of manual effort.

The Physical AI team builds simulation, synthetic data, and validation tooling that enables autonomous systems to be tested systematically and safely before deployment. We work across the full Physical AI lifecycle — data curation, scenario generation, simulation, and deployment monitoring. Our team collaborates closely with Applied AI Solutions teams (IoT, Digital Twin, Spatial/Geospatial), SageMaker teams for model development, and leading simulation platform providers. We value technical excellence, customer obsession, and systematic problem-solving. You'll have opportunities to shape the future of autonomous systems validation, work with generative AI technologies, and deliver measurable impact to customers deploying safety-critical autonomous systems at scale.


Key job responsibilities
Design and implement scenario generation pipelines that transform operational design specifications into diverse scenario variations with comprehensive coverage analysis. Build generative AI model integration layers that leverage real-world operational data to create realistic behaviors while maintaining physical plausibility constraints for agent dynamics, sensor characteristics, and environmental physics. Develop export connectors for industry-standard simulation platforms that handle format compatibility, authentication, and data transfer.

Create synthetic sensor data generation systems that produce multi-modal outputs (MP4 video, PCD/LAS point clouds, radar data) with accurate sensor characteristics and metadata tracking for validation traceability. Implement coverage analysis algorithms that identify gaps in generated scenario distributions and recommend generation parameters to achieve systematic coverage. Integrate with SageMaker for perception model training workflows and visualization tooling for validation coverage reporting.

A day in the life
You'll start your morning reviewing generation quality metrics from customer validation campaigns, analyzing scenario success rates and identifying opportunities to improve edge case coverage. You'll participate in a design review for the next iteration of a specification parser, discussing how to handle region-specific rules and regulatory requirements. Mid-morning, you'll pair program with a teammate on implementing physics-based validation checks that verify generated scenarios meet dynamics constraints and behavioral plausibility. After lunch, you'll join a customer call to understand their specific validation challenges and gather requirements for specialized sensor modalities. You'll spend the afternoon optimizing a generation pipeline for scalability, balancing generation quality with compute costs for large scenario volumes. You'll end the day mentoring a junior engineer on generative AI model integration patterns and reviewing pull requests for the simulation platform connector framework.
About the team


About the team
ABOUT AWS:
Diverse Experiences
Amazon values diverse experiences. Even if you do not meet all of the pre