CodeNext - Software Engineer

Who is Sonar?

Sonar helps prevent code quality and code security issues from reaching production, amplifies developers' productivity in concert with AI assistants, and improves the developer experience with streamlined workflows. Sonar analyzes all code, regardless of who writes it — your internal team, genAI, or third parties — resulting in more secure, reliable, and maintainable applications. Rooted in the open source community, Sonar’s solutions support over 30 programming languages, frameworks, and infrastructure technologies.

Today, Sonar is used by +7M developers and 400K organizations worldwide, including the DoD, Microsoft, NASA, Master Card, Siemens, and T‑Mobile.

We believe in developing great products that are supported by great internal teams and a strong culture. We are highly committed to and obsessed with the company, users, each other, and our open source community. We have high standards and hold each other accountable for acting with positivity, dedication, thoughtfulness, empathy, and passion daily.

We are deliberate with our decisions with high clarity of intention. At the same time, we feel extreme urgency and move forward quickly.

And lastly, we are highly effective and operationally efficient. We operate collectively as One Team to accomplish our goals.

At Sonar, CODE is more than just an acronym – it's a mindset that defines daily operations.

At Sonar, we’re a group of brilliant, motivated, and driven professionals working hard to help supercharge developers to build better, faster. Sonar helps to continuously improve code quality and code security while reducing developer toil. This means that developers can focus on doing more of what they love and less of what they don’t. Our solutions don’t just solve symptoms of problems – we help fix issues at the source – for all code, whether it's developer‑written, AI‑generated, or from third parties.

We have a dynamic culture with employees worldwide and hub offices in the USA, Switzerland, the UK, Singapore, and Germany. Team members should be able to come to work every day, work on a product they are proud of, love what they do, and feel energized by their peers. With our roots deep in the open source community, we’re all about the mission: supercharge developers to build better, faster.

The Sonar Code Next team is looking for a visionary Staff Software Engineer to join a fast‑paced, dynamic group at the intersection of Large Language Models (LLMs), static analysis, and programming languages. As a technical leader in this team, you’ll directly influence AI‑driven solutions by designing and scaling the tools and infrastructure that make these initiatives possible. In this hands‑on, high‑impact role, you’ll focus on architecting methodologies that enhance code quality—ensuring that all solutions are scalable, maintainable, and ready for global production.

Working on cutting‑edge, AI‑driven projects, you’ll architect and steer the boundaries of automated code quality and security.

Our mission is to harness our expertise in static analysis and LLMs to advance AI’s ability to understand, generate, and improve code. We develop prototypes grounded in cutting‑edge research, testing innovations across the LLM development lifecycle—from data processing and model training to benchmarking. Our approach leverages techniques such as supervised fine‑tuning (SFT), direct preference optimization (DPO), reinforcement learning from human feedback (RLHF), and retrieval‑augmented generation (RAG) to enhance model performance and code quality.

• Architect and oversee the tooling and infrastructure to support AI‑driven solutions for code quality and security across diverse language ecosystems.
• Lead the design and optimization of AI‑focused experiments, creating robust systems to evaluate modern techniques, including fine‑tuning, RAG, DPO, and RLHF.
• Write scalable, production‑quality code mainly in Python, and sometimes in Java, Type Script, or Rust, to support AI workflows, dataset processing, and evaluation frameworks.
• Design and maintain high‑throughput benchmarking pipelines to assess LLM…