Want to know if this job is worth applying to?
San Francisco, California, United States
Remote only
CodeRabbit is the governance and control layer for agentic software development, helping engineering teams trust the code they ship. AI can generate code faster than people can review and understand it. Teams still need to know whether each change solves the right problem, fits their architecture, and is safe to ship.
We provide independent validation grounded in codebase context and team standards, giving people the evidence they need to judge what should ship. Across Review, Triage, Explain, and Secure, our platform challenges changes before merge, prioritizes pull requests by value, impact, and risk, explains what changed and why it matters, and monitors the codebase for security vulnerabilities after merge.
CodeRabbit serves nearly 20,000 customers and supports approximately 300,000 open-source repositories, performing over 2 million code reviews each week. Our $143M Series C, raised at a $1.5B valuation, supports further product development and international expansion.
As an Applied Gen AI Engineer at CodeRabbit, you'll play a central role in designing, building, and deploying advanced generative AI systems that power our code review and developer productivity tools. You’ll be responsible for bringing the latest advancements in generative AI to life — integrating techniques like RAG, RLHF, and multi-step agentic reasoning into high-impact product workflows.
You’ll collaborate with engineers, product managers, and technical leads to iterate on intelligent systems that deliver real-world value, improving how developers write, review, and ship code.
Design and optimize LLM-based systems for high-quality, context-rich code reviews
Build and refine agentic workflows that reason across multiple steps and contexts
Develop and maintain knowledge base and retrieval pipelines (e.g., chunking, embeddings, semantic search)
Deploy generative AI models and pipelines into production and monitor performance
Collaborate across teams to ensure that AI outputs align with user needs and product goals
Analyze human-in-the-loop feedback and usage data to iteratively improve system performance
Apply RLHF, ranking, and reward modeling techniques to improve response quality over time
Stay current with the latest generative AI developments and apply them to new use cases
Education: Degree in Computer Science, Engineering, Artificial Intelligence, or related field, or equivalent practical experience
Experience: 5+ years applying ML or LLM-based systems in real-world production environments, with at least 2 years of industry experience focused on generative AI
Technical Skills: Strong programming skills in TypeScript and Python
AI Frameworks: Experience with tooling such as LangChain, LlamaIndex, OpenAI APIs, or vector databases like Pinecone or Lancedb
Prompt Engineering: Strong skills in prompt engineering
Data Fluency: Ability to extract insight from telemetry, logs, user signals, and structured feedback
Practical Mindset: Comfortable applying research-inspired methods to solve concrete product challenges
Cross-Functional Collaboration: Experience working across product, engineering, and design to deliver production-grade systems
Experience optimizing RAG systems and tuning retrieval performance using custom embeddings or search strategies
Hands-on experience with RLHF pipelines, reward modeling, or behavioral policy tuning in LLMs
Experience integrating LLM systems into developer tooling or collaborative workflows
Track record of contributions to open-source projects or publications in applied AI/ML
CodeRabbit is building the next generation of AI-native developer tooling — starting with code review. We combine large language models with deep software engineering context to help teams ship faster, catch more bugs, and make better architectural decisions at scale.
We are a high-ownership engineering culture. That means no passive execution, no waiting for perfect tickets, and no narrowly defined task boundaries. Engineers here find problems before they're assigned, use AI as a core part of how they build, ship with judgment, and own outcomes from proposal to production.
Our operating philosophy: bias toward action, ship the smallest necessary coherent slice, validate proportional to risk, watch what happens, and make the system better. AI drafts; humans decide. Speed matters, but so does understanding what you ship.
This opportunity will be energizing for people who want real ownership, pace, and high standards. It's uncomfortable for people who prefer slow consensus or heavily managed workflows.
If you want to build tools that are changing how software gets written, and be held to the standard that the best engineers thrive under; we'd love to talk.
🤝 Collaborative Humans — Prioritizing collective intelligence
🚀 Fearless Innovators — Turning obstacles into growth opportunities
💪 Persistent, Passionate Developers — Thriving on complex, long-term challenges
🎯 Impact-Driven Creators — Crafting intuitive tools for developers
🧠 Rapid Learners and Un-learners — Adapting quickly in our fast-paced technological world
