About the company
At Databricks, we are passionate about enabling data teams to solve the world's toughest problems — from making the next mode of transportation a reality to accelerating the development of medical breakthroughs. We do this by building and running the world's best Data Intelligence Platform so our customers can use deep data insights to improve their business.
Responsibilities
- Evaluate ML and LLM approaches for CustomerLake's personalization use cases, push the models and algorithms forward, and continuously improve quality over time.
- Go deep on how models behave in production: inspect individual traces, understand how the models reason, and tune and improve from there.
- Build the platform and evaluation framework that let CustomerLake customers optimize for real business value such as purchases, retention, and product usage, not vanity metrics like email opens and clicks.
- Push the team toward new directions and novel methods worth tackling, not just optimizing what already exists.
- Partner closely with product management, engineering, and design to turn ambiguous customer problems into scalable, trustworthy solutions.
- Set the technical foundation and best practices for our ML/AI personalization work as we grow this into several roles across our products over the next 1-2 years.
Requirements
- 10+ years of engineering experience, with a strong foundation across the full loop of shipping and improving ML/AI products.
- Hands-on experience building and evaluating ML models and/or LLM systems for real product or business use cases.
- Experience with personalization based on customer behavior (ideal) or transactions (acceptable), such as recommendations, targeting, churn, or lifetime-value modeling.
- Proficiency in Python and modern ML frameworks (e.g., PyTorch), with hands-on experience in model evaluation and monitoring AI quality in production.
- Familiarity with LLMs and generative AI, including techniques like retrieval-augmented generation (RAG), prompt design, fine-tuning, and evaluation.
- A demonstrated product mindset, with the ability to translate ambiguous customer problems into scrappy MVPs and iterate quickly based on data and user feedback.
- High ownership and bias for action in 0-to-1 environments: comfortable making pragmatic trade-offs, operating with incomplete information, and driving projects from idea through launch and adoption.