Software Engineer (Java) - Remediation Agent

Sonar•Geneva, IL

12d•Onsite

About The Position

At Sonar, we believe that simply reporting code issues is not sufficient. To truly level up the quality of the world's software, we must empower millions of developers to fix them automatically. With over 400,000 enterprises relying on our products, we hold a unique lever to shape how software is built. Recent breakthroughs in AI have made autonomous remediation a reality, but the industry faces a major hurdle: Trust. Enterprises cannot automate what they do not trust. As a Software Engineer in the CodeGen squad, you will build the critical backend components for Verified Autonomous Remediation. You will work within a high-performing team to turn research prototypes into reliable, production-grade features that test and validate AI-generated code. You will help us close the "Trust Gap" by delivering robust software that just works.

Requirements

Solid Backend Engineering (3+ years). You have a track record of building and maintaining backend systems. You write clean, testable code and care about software craftsmanship.
Polyglot Versatility (Java & Python). You are comfortable with one of the core languages: Java (for integration with the Core Sonar ecosystem) or Python (for AI/Agent logic). You pick up new tools and syntax quickly.
Cloud Operations (AWS). You understand how your code runs in the cloud. You have experience with AWS (Lambda, ECS, or similar) and are comfortable with deployment pipelines, logging, and basic infrastructure monitoring.
Systems Fundamentals. You understand how different services talk to each other (REST, queues, async processing). You know how to debug a distributed system when things go wrong.
Interest in AI/LLMs. You are excited about the potential of AI in software development. Experience using LLM APIs (OpenAI/Anthropic) is a plus, but a willingness to learn how to integrate them into backend workflows is essential.
Ownership Mindset. You don't just "take tickets." You own the delivery of your tasks, raising risks when necessary and ensuring your work meets high quality standards before it reaches production.

Nice To Haves

Experience using LLM APIs (OpenAI/Anthropic) is a plus

Responsibilities

Build Features for the "Unified Remediation Engine". You will implement the backend logic that powers our remediation efforts. You will write clean, reliable code (Java/Python) that routes requests between deterministic QuickFixes and LLM agents, ensuring features are delivered reliably within our existing framework.
Implement Production-Grade Agent Logic. You will take concepts from the Research team and "productionize" them. You will write the specific workflows that allow the Agent to read CI/CD errors, attempt fixes, and iterate until the build is green.
Own Your Code from Local to Cloud. You are responsible for the full lifecycle of the features you build. You will deploy your services to AWS, monitor their health (metrics/logs), and troubleshoot production issues to ensure high availability for our users.
Secure & Optimize Execution. Working within our existing sandboxing framework, you will implement the runners and isolation layers required to execute AI-generated code safely. You will ensure your implementation is secure and efficient.
Collaborate & Share Knowledge. You will be an active member of the squad, reviewing PRs, documenting your work, and sharing knowledge with peers. You will act as a subject matter expert for the specific components you own, ensuring the team moves forward together.