Staff AI/ML Engineer

View full description

Requirements:

• 5+ years of experience in ML/AI, including shipping models to production and iterative improvement post-launch.
• Strong knowledge of LLMs, including SOTA models (GPT, Claude, Mistral, etc.), with practical experience in prompt engineering, fine-tuning, or retrieval-augmented generation.
• Deep understanding of NLP, including tokenization, embeddings, and transformer architectures.
• Deep understanding of machine learning, including experience with some fields of classical ML (recommendation systems, regressions/classifications on tabular data, and time series or other areas of classical ML).
• Ability to work with customer data to identify usage patterns, perform analytics, and generate insights for product development and model optimization.
• Ability to set up data collection pipelines.
• Solid understanding of software engineering concepts, especially around the SDLC in modern dev environments.
• Proficient in Python and ML frameworks (PyTorch, HuggingFace, etc.).
• Proven ability to work effectively in a collaborative team environment, with excellent communication skills and a commitment to delivering high-quality solutions on time.
• Experience designing and evaluating AI agents or multi-agent pipelines is a strong plus.

Responsibilities:

• Design, build, and optimize LLM-powered agents that assist developers in tasks such as code generation, unit test creation, bug fixing, and refactoring.
• Research the capabilities and limitations of SOTA LLMs and apply findings to improve agent performance and reliability.
• Develop evaluation pipelines to benchmark model quality, correctness, and impact on developer productivity.
• Collaborate cross-functionally with product, software, and infrastructure teams to integrate AI agents seamlessly into IDE environments (JetBrains, VSCode).
• Craft effective prompts, fine-tune models, and experiment with advanced techniques like RLHF or DPO to guide model behavior.
• Analyze user interactions and data to derive insights and continuously optimize the agent experience.