Software Development Engineer , Amazon Robotics (AR) Sortation Planning

Amazon•North Reading, MA

About The Position

Are you motivated by building systems that solve complex, real-world problems at scale? Do you enjoy working with teams to turn ideas into production systems that have measurable impact? At Amazon Robotics, we are a team of builders applying advances in robotics, software, and AI to transform how fulfillment networks operate. Our work directly shapes the efficiency, safety, and reliability of systems used across Amazon’s global operations. Are you excited about building systems that don’t just predict—but reason and take action in real-world environments? The AR Sortation Insights team develops intelligent platforms that power next-generation warehouse operations across three key areas: Simulation – modeling and optimizing robotic workflows ML Ops – scalable data pipelines, model training, and inference systems ARORA (Amazon Robotics Outbound Remote Assistant) – agentic AI systems that observe, reason, and act in real time This role is primarily focused on ARORA, where we are building agentic AI systems that continuously monitor operational signals, generate insights, and safely trigger actions to optimize robotic workflows at scale. As a Software Development Engineer, you will design and build closed-loop intelligent systems that combine machine learning, real-time data processing, simulation, and large language models (LLMs). Your work will directly influence how Amazon’s robotic systems operate—driving improvements in throughput, safety, and efficiency across global fulfillment networks.

Requirements

5+ years of non-internship professional software development experience
2+ years of non-internship design or architecture (design patterns, reliability and scaling) of new and existing systems experience
2+ years of full stack development experience
Experience building complex software systems that have been successfully delivered to customers
Experience with full software development life cycle, including coding standards, code reviews, source control management, build processes, testing, and operations

Nice To Haves

Master's degree in computer science, machine learning, engineering, or related fields
5+ years of software development experience
Strong programming skills in Python and/or Java, and proficiency with AI frameworks like ( Strands,Langchain), AWS Bedrock,Langfuse and went through E2E Agentic AI development
Solid understanding of machine learning algorithms, model evaluation, and data-driven optimization.
Experience building and deploying ML models in production environments.
Comfort working in a fast-paced, high-impact team environment with end-to-end ownership.
Experience building and maintaining applications with JVM-based languages, such as Java, Kotlin, Scala.
Knowledge of professional software engineering & best practices for full software development life cycle, including coding standards, software architectures, code reviews, source control management, continuous deployments, testing, and operational excellence

Responsibilities

Design and build agentic AI systems that combine machine learning models, real-time data, and LLM-driven reasoning to automate operational decision-making in robotic workflows.
Develop end-to-end decision systems that observe system state, generate insights, and trigger safe, validated actions through robust safeguards and control mechanisms.
Build scalable, event-driven architectures on AWS to support real-time and batch processing for inference, decisioning, and continuous feedback loops.
Design and implement machine learning models for perception, classification, prediction, and optimization, and integrate them into production systems.
Develop and manage data pipelines, model training workflows, and deployment infrastructure, ensuring reliable and scalable ML Ops practices.
Collaborate with simulation teams to test and validate decision strategies in virtual environments prior to production rollout.
Conduct offline and in-simulation experimentation and analysis to continuously improve model performance and system outcomes.
Partner with cross-functional teams including data scientists, data engineers, simulation developers, and operations leaders to deliver end-to-end solutions.
Implement observability, monitoring, and evaluation frameworks to ensure reliability, safety, and performance of automated systems.
Own production systems end-to-end, including deployment, monitoring, retraining, and continuous improvement.
Contribute to the technical architecture and roadmap for agentic AI and ML-driven robotics platforms within AR Sortation Insights.