Senior Agentic Software EngineerThe RoleThe Senior Agentic Software Engineer delivers production features, modernizations, and AI-powered capabilities across our stack. This is an AI-native role. The expectation is that agentic workflows are your default mode of working, from picking up a requirement through delivery, not a supplement to your existing process.
You will work alongside our Staff Agentic Engineering Lead and pair with existing engineers to transfer agentic working habits through real, shared work on production problems. You will contribute to sprint commitments with consistency and take genuine ownership of the systems you build, including when they break.
What You Will Do- Deliver production features and modernizations across our stack using agentic development as your primary workflow
- Work in brownfield environments, adding new functionality to existing code with the same AI-native discipline as greenfield work
- Build out and operate multi-agent workflows where distinct roles (requirements, architecture, testing, development) operate in parallel and in coordination
- Pair with existing engineers on real problems to transfer agentic working habits in practice
- Validate agent output continuously, knowing when to intervene and when to let work run based on how well requirements were established upfront
- Contribute to timely sprint delivery and consistently meet or exceed release schedule targets
- Maintain software quality by managing escaped defects and contributing to backlog reduction over time
- Produce scalable, efficient, and well-documented code
What We Are Looking ForEngineering Background- 5 to 9 years of professional software engineering experience
- Strong fluency in Java or Python on AWS, or C# on Azure
- Production ownership history: you have shipped non-trivial systems and been accountable when things went wrong
- At least one significant out-of-depth experience, whether a stack switch, a domain change, or a greenfield build in unfamiliar territory
Agentic Development Experience- 6 or more months of serious, production-grade agentic tool use (Claude Code, Cursor, or equivalent). Production work only, not pilots or experiments
- Demonstrated ability to build product with AI, not just use AI within existing workflows
- Concrete understanding of when to let an agent run versus when to intervene, grounded in how well requirements were established before the work started
- Experience reviewing and validating work you did not write, including agent-generated output
- Familiarity with building multi-agent frameworks where distinct roles operate concurrently
