Zillow is looking to streamline the loan production process to enable a fully digital financing experience and provide consistent, unbiased, and personalized loan recommendations, making the home-buying process less daunting and more manageable.
Requirements
- Hands-on AI practitioner. 3+ years building and assisting in the production of AI/ML systems in Python with frameworks such as PyTorch, TensorFlow, or Scikit-learn.
- Agentic + GenAI skills. Familiar with LLMs, multi-agent systems, tool-augmented reasoning, or other GenAI methods; excited to apply them in production.
- Traditional ML foundation. Comfortable with predictive modeling and data-driven experimentation when reliability and scale are key.
- Data wrangler. Experienced with SQL and distributed tooling (Spark, Hive, Databricks) for large-scale, messy data.
- Engineering mindset. Strong coding practices, including Git, CI/CD, testing, and code reviews.
Responsibilities
- Develop agentic workflows. Build intelligent agents that use LLMs, planning, and tool integration to support Loan Officers and improve borrower experiences.
- Prototype and validate. Rapidly design and test solutions with offline evaluations and online experiments, and ship the winning approaches.
- Blend approaches. Combine generative reasoning with classical ML (e.g., scoring, ranking, recommendation) for end-to-end decisioning.
- Apply cutting-edge research. Track advances in GenAI and ML, bring relevant techniques into practice, and share insights across the org.
- Communicate across functions. Clearly explain findings and trade-offs to scientists, engineers, and business partners to drive alignment.
Other
- Collaborative communicator. Able to translate technical depth for varied audiences and build trust across disciplines.
- 3+ years of experience
- U.S. employees may live in any of the 50 United States, with limited exceptions.
- Must be able to work remotely
- Must be willing to sign a Proprietary Rights Agreement
