Senior Software Engineer - Agentic First

Sonatype is the software supply chain security company that provides leading solutions for secure software development. As a Senior Software Engineer, you will design and implement agentic-first features within Nexus Repository Manager, focusing on multi-agent development workflows to enhance software supply chain security.

Responsibilities:

• Design and implement scalable, high-performance features across the stack by driving long-running, multi-agent development workflows end-to-end — decomposition, orchestration, implementation, testing, and review
• Take features from ambiguous requirements through design, implementation, rollout, and on-call operability, using agents to move faster without compromising quality
• Spend your time on direction, review, evals, and testing rather than line-by-line coding. Build and use the harnesses and guardrails that let you trust what the agents output
• Contribute to internal playbooks, tooling, and rituals for how Sonatype engineers work with agents — sharing what's working, what isn't, and what we should try next
• Write and ship clean, well-tested, observable code. Apply software supply chain security best practices to everything you deliver
• Partner with Product, UX, and fellow engineers to translate customer needs into shippable solutions, and raise the quality bar through thoughtful code review and mentorship of earlier-career engineers

Requirements:

• 4+ years of professional software development, including meaningful experience shipping and operating production services
• Strong experience with Java and at least one major cloud (AWS / Azure / GCP). Comfortable working in distributed systems — APIs, databases, queues, and the failure modes that come with them
• Familiarity with concepts like SBOMs, SCA, vulnerability management, dependency hygiene, and artifact/package ecosystems — or strong interest in going deep here
• Multi-step, long-running agent workflows are your default way of building software — well beyond Copilot-style autocomplete
• Hands-on experience running and composing multiple agents (e.g., Claude Code, Codex, Cursor background agents, or equivalents) — including MCP tools, custom prompts/skills, shared context, and eval loops that keep output trustworthy
• You've internalized that the new leverage point is human judgment over machine generation. You write and maintain evals, test harnesses, and review workflows that let you confidently ship code you didn't personally type
• You're energized by being early in a new way of building software, tracking what's happening on the frontier, and bringing new techniques back to your team
• You think in terms of customer outcomes, not just tickets, and can make sensible product trade-offs in partnership with PM and design
• You want to build mission-critical products that drive revenue and transform how customers build software
• You push agentic tools to their limits — probing where they work, where they break, and how to make them better. You're energized by being early in a fundamentally new way of building software