Agentic Software Engineer

Agentic Pods is building a new model for how Engineering partners with teams across Uber to transform work with AI. Our engineers embed directly with operational teams to understand complex workflows, reimagine how that work could operate in an agentic-first world, and build the production systems that make that future real.

We are looking for an Agentic Software Engineer to design, build, and operate production-grade agentic systems and processes that improve complex operational workflows.

This role combines backend engineering, reliability engineering, and a forward-deployed working model. You will work directly with operators and domain experts to discover opportunities, determine where agents can safely help, and turn ambiguous problems into reliable software.

You will own solutions from discovery through architecture, implementation, production rollout, and ongoing operation.

Discover and Build Agentic Systems

• Embed with operational teams to understand workflows and identify where agentic AI, deterministic software, automation, or human decision-making can meaningfully improve how work gets done.
• Design, build, and iterate on production agentic systems, including agents, skills, backend services, APIs, integrations, data pipelines, and human-in-the-loop workflows, alongside real users.

Own Production Reliability

• Define reliability and quality targets and build observability across agent decisions, model behavior, dependencies, and workflow outcomes through tracing, logging, metrics, alerting, fallbacks, and recovery.
• Own systems in production, using incidents, operational data, and user outcomes to continuously improve performance, resilience, and reliability.

Make Agents Safe and Measurable

• Define evaluations and validation approaches for task completion, output quality, tool selection, policy compliance, and failure handling across offline testing, controlled rollouts, and production monitoring.
• Build appropriate controls—including permissions, approvals, auditability, data minimization, reversible execution, and human escalation—and partner with Security, Privacy, Legal, and Data to ensure consequential actions are safe, explainable, and recoverable.

Build for Reuse and Scale

• Identify capabilities that should become reusable skills, shared services, platform components, or standard integration patterns, supporting business and regional differences through configuration rather than unnecessary system forks.
• Establish common contracts and contribute to architecture, tooling, deployment patterns, reliability practices, and technical standards that enable Agentic Pods to build and scale consistently.

Who Thrives on This Team

We're looking for engineers with strong technical depth who are comfortable operating in ambiguity and working directly with users.

• Agentic-first thinker: You naturally look for opportunities to redesign workflows around what's newly possible.
• 0→1 builder: You move quickly from ambiguous problems to working software.
• Product-minded: You have strong judgment about what is useful and worth building, not just what is technically possible.
• Systems thinker: You recognize when individual solutions should become reusable capabilities or standards.
• Adaptable communicator: You learn unfamiliar domains quickly, build trust, and communicate effectively with technical and non-technical partners.

Minimum Qualifications

• Bachelor's degree or equivalent in Computer Science, Engineering, Mathematics or related field OR 1-year full-time Software Engineering work experience, WHICH INCLUDES 1-year total technical software engineering experience in one or more of the following areas:
• Programming language (e.g., C, C++, Java, Python, or Go).
• Note: The 1-year total of specialized software engineering experience may have been gained through education and full-time work experience, additional training, coursework, research, or similar (or some combination of these). The year of specialized experience is not necessarily in addition to the years of education and full-time work experience indicated.
• Strong experience designing and building production backend or distributed systems, including agentic applications, AI-enabled workflows, orchestration systems, developer platforms, APIs, asynchronous workflows, and system integrations.
• Experience owning and improving production systems, including monitoring, alerting, incident response, failure recovery, and debugging across application, infrastructure, and data dependencies.

Preferred Qualifications

• Experience integrating LLMs with tools, APIs, enterprise data, and permissioned systems.
• Familiarity with agent evaluation, retrieval, context management, model routing, and safety controls.
• Experience in SRE, infrastructure, platform, production engineering, or a forward-deployed/embedded engineering model.
• Experience designing systems for sensitive or operationally consequential workflows.
• Ability to lead technical design, drive key architectural decisions, and influence technical direction across teams.

Deep AI research experience is not required. We value engineers who can apply emerging agent technologies within secure, reliable production systems.