Founding AI engineer

At nao Labs, we are reimagining how data work gets done. Data work is different from traditional software development, and it deserves a specialised workflow. That’s why we’re creating nao—an AI-powered code editor tailored for data workers. We're looking for a Founding AI Engineer to join nao Labs and help define the future of data work. You should have deep experience across the generative AI stack—from model training, fine-tuning, and evaluation to advanced techniques like pre-prompting, RL for Fill-in-the-Middle (FIM), code generation, and task-specific adaptation. You will: Lead AI development at nao Labs Train, fine-tune, and evaluate models for data engineering, analytics and science related tasks Apply RL and other techniques for next-gen FIM and codegen systems Build internal tooling for efficient experimentation and deployment Lead AI development at nao Labs Train, fine-tune, and evaluate models for data engineering, analytics and science related tasks Apply RL and other techniques for next-gen FIM and codegen systems Build internal tooling for efficient experimentation and deployment You are: An expert in LLMs, transformers, RLHF, and code models Hands-on with training pipelines, distributed training, and evaluation metrics Entrepreneurial — ready to build, ship, and iterate fast An expert in LLMs, transformers, RLHF, and code models Hands-on with training pipelines, distributed training, and evaluation metrics Entrepreneurial — ready to build, ship, and iterate fast nao is a fork of VS Code, which means NodeJS and Typescript. We also use React for webviews. Python/Django and NodeJS for our backend APIs. A lot of LLMs to power our product. The biggest challenges are: Navigating VS Code codebase to build a lovable product for data people Understand SQL and Python codebase to provide syntaxic trees Build low-latency AI Tab features and agentic workflows Navigating VS Code codebase to build a lovable product for data people Understand SQL and Python codebase to provide syntaxic trees Build low-latency AI Tab features and agentic workflows Async screening through a list of technical and non-technical questions 30’ phone call to asses motivation and profile matching 1h30 technical pair-programming session to evaluate technical level with CTO and a member of the technical team 30’ with the CEO Async screening through a list of technical and non-technical questions 30’ phone call to asses motivation and profile matching 1h30 technical pair-programming session to evaluate technical level with CTO and a member of the technical team 30’ with the CEO