Sonar is the independent AI code verification platform trusted by 7M+ developers & organizations to help build secure, reliable software with confidence. Backed by General Catalyst and Insight.
About the role
Spend the majority of your time (roughly 70%) designing, building, and iterating on agentic tooling: orchestration, tool-use, context management, and evaluation harnesses for autonomous code-modification agents. Build and refine the verification layer that gates what agents are allowed to ship — automated test generation, regression detection, static analysis (using our own SonarQube), and human-in-the-loop checkpoints where warranted.
What they're looking for
- Hands-on experience building or operating multi-agent or LLM-agent systems in a real production setting — not just prompting a coding assistant, but designing the orchestration, tool-use, and guardrails around autonomous agents
- Experience building evaluation or verification systems for AI-generated changes: automated testing, static analysis, regression detection, or comparable safety nets for code you didn't personally write
- Strong software engineering fundamentals in Java and Spring (or a comparable JVM/enterprise stack) — enough to read a large, mature codebase critically and tell good architecture from bad
- A track record of owning complex, ambiguous technical problems independently and driving them to resolution — this is an R&D-flavored role inside a production platform team, and you'll be shaping the playbook, not following one
- Experience influencing engineers and squads you don't manage — the migrations you'll enable depend on their trust as much as on your code
More about this role
Sonar is driving the future of agent-centric software development. As the leader in AI code review and verification, we solve a critical problem: ensuring that software generated by AI-assisted developers or autonomous agents is reliable, secure, and maintainable.
Integrating seamlessly with Claude Code, Codex, Cursor, GitHub Copilot, Gemini, and Devin, we help over 75% of the Fortune 100 build trusted, reliable, compliant software. Customers who use Sonar are 44% less likely to report an outage due to AI-generated code.
We believe code verification is the critical missing link in the Agent-Centric Development Cycle (AC/DC). Industry giants like Nvidia, ServiceNow, Booking.com, Goldman Sachs, AstraZeneca, and Ford Motor Company count on us to provide independent, explainable, consistent review and governance of their AI-generated code via products like:
SonarQube: The world's leading AI code review and verification platform.
SonarQube Foundation Agent: Currently topping the leaderboards for agentic software repair.
SonarSweep & Sonar Context Augmentation: Providing the enterprise-grade context and constraints agents need to be truly effective.
Our team operates across global hubs...
Browse similar: Startup jobs