AI Software Engineer

This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for an AI Software Engineer based in United States. This is a full-stack engineering role focused on building and scaling production-grade agentic AI systems for complex physical operations. You will work across the entire product, from React and TypeScript interfaces to agent orchestration, Python services, and AWS infrastructure. The platform is already deployed across hundreds of demanding warehouse and manufacturing environments, creating a strong balance between innovation and production engineering. You will build new AI capabilities while improving the reliability, latency, safety, and extensibility of systems already serving enterprise customers. The role involves integrating frontier models from multiple providers rather than training models from scratch, with an emphasis on engineering the systems around them. You will work closely with real operational users, using feedback and measurable outcomes to continuously refine AI-powered workflows. This is an ideal opportunity for a hands-on engineer who enjoys solving ambiguous problems across the stack and turning sophisticated AI capabilities into dependable products. Build and ship full-stack product features spanning user interfaces, APIs, backend services, and agent logic. Develop and extend agent frameworks supporting tool and function calling, multi-step reasoning, context management, and reliable task execution. Create structured and secure access to operational data through schema grounding, validation, safe execution, and integrations with external systems. Build evaluation and observability capabilities to measure AI quality, latency, safety, cost, reliability, and product outcomes such as task completion. Improve the reliability, robustness, performance, and latency of distributed, event-driven backend systems. Re-architect and modernize early implementations as the platform scales and production requirements evolve. Design and maintain appropriate AI guardrails, including least-privilege tools, safe execution boundaries, fallbacks, and emergency kill switches. Contribute across the technology stack as required, including React and TypeScript interfaces, Electron desktop applications, streaming experiences, and Python services running on AWS. Build interfaces that remain responsive and useful while AI-generated output is still being produced. Integrate and orchestrate large language models from multiple providers to deliver reliable, user-facing AI functionality. Use operator and customer feedback to continuously improve AI experiences and handle uncertain or non-deterministic model output appropriately. Apply agentic coding tools such as Claude Code, Codex, Cursor, or comparable technologies to accelerate development and engineering productivity. Requirements 5+ years of experience building production full-stack or backend software, or equivalent demonstrated technical depth. Recent hands-on experience integrating large language models into production, user-facing applications. Proven experience shipping an agentic or LLM-powered product feature and owning it across the full technology stack. Experience building evaluation harnesses and quality measurement systems for non-deterministic AI applications. Demonstrated track record of improving the reliability, performance, scalability, and maintainability of production systems. Strong hands-on proficiency in both Python and TypeScript/React, with experience using both regularly in production environments. Experience building agent orchestration systems or deep familiarity with an agent framework and its technical trade-offs. Strong understanding of REST API and service design, along with asynchronous and event-driven architectures such as queues, streaming, and WebSockets. Production experience deploying, operating, and troubleshooting services on AWS. Strong product judgment for user-facin