Define and execute the applied research roadmap for the Large User Model ecosystem, balancing immediate customer needs with long-term technical evolution to scale foundational models across high-traffic surfaces.
Support initiatives with product leadership to translate complex business goals into technical model formulations, delivering step-function improvements in user engagement and business metrics while optimizing for latency and performance trade-offs.
Optimize model performance by researching and implementing adaptation techniques (transfer learning/domain adaptation) that balances high-quality output with strict inference latency requirements for production environments.
Drive architectural improvements by establishing a strategic feedback loop with pre-training teams, utilizing downstream performance analysis to influence data curation, model architecture, and novel evaluation metrics for engagement-specific needs.
Design and productionize fine-tuning pipelines that translate general-purpose state-of-the-art (SoTA) Foundation User Models into effective, domain-specific recommendation engines.

Staff Software Engineer, Applied Research, Foundation User Models

Job Description