Principal AI/ML Researcher / Engineer In Bayesian, Large Foundational Systems, and Distributional Reinforcement Learning

Airbnb was born in 2007 when two hosts welcomed three guests to their San Francisco home, and has since grown to over 5 million hosts who have welcomed over 2 billion guest arrivals in almost every country across the globe. Every day, hosts offer unique stays and experiences that make it possible for guests to connect with communities in a more authentic way. About the Role We are seeking a seasoned Principal AI/ML Researcher and Engineer with deep expertise in Bayesian Learning, and Distributional Reinforcement Learning (RL) to lead the advanced research and development of cutting-edge intelligence AI models. These systems will integrate foundational Bayesian frameworks with advanced architectures, including Mixture of Models, multi-pass sharded systems, multitask and multi-objective optimization, and external knowledge incorporation. Additionally, the role involves innovating ways to interoperate and integrate Large Language Models (LLMs) and Large Multimodal Models (LMMs) with Reasoning, Planning, and Decisioning abilities into the Bayesian frameworks to create a seamless foundational model fabric that synergizes with diverse model ecosystems.The role will require ensuring these models and supporting systems perform efficiently at scale, integrating them into live systems that directly impact product and user experience. Our goal is to build next-generation AI platforms that redefine personalization, decision-making, and intelligence across diverse applications. You will work on developing production-level systems, collaborate with cross-functional teams, and play a pivotal role in shaping our AI/ML strategy. Relevance and Impact of This Role This role drives Airbnb's evolution toward probabilistic, uncertainty-aware intelligence systems capable of reasoning under ambiguity and learning continuously from dynamic environments. The near-term impact spans improved personalization quality, ranking quality, uncertainty estimation, and adaptive decision-making across guest and host experiences — enabling policy-driven intelligence that handles long-tail discovery, evolving preferences, and complex marketplace dynamics. Longer term, this role helps establish Airbnb's leadership in adaptive probabilistic intelligence by building the foundational substrate that connects Bayesian learning, reinforcement learning, foundational models, multi-agent orchestration, and large-scale personalization into a unified adaptive architecture — where AI systems continuously balance exploration, exploitation, uncertainty, and ecosystem optimization at scale. What You Will Do Research & Innovation: Lead groundbreaking applied research in Bayesian systems, distributional reinforcement learning, and multi-modal architectures to drive novel advances in AI and Foundational Intelligence (Ranking, Recommendations, Personalization) to fill out gaps in the Long Tail Curve of Discovery in order to grow the Business Offerings on both Guest and Host Long Tail Ends Bridge the gap between theoretical AI/ML advancements and real-world production systems Ensure that new research can be effectively applied and scaled to meet practical needs. Architect and Design : Define and drive the architecture of large-scale Bayesian Framework-based AI systems at Airbnb. Develop multi-pass sharded Bayesian + Discriminative/Generative single to multi agent systems for scale and efficiency. Incorporate Mixture of Models and Agents, multitask learning, multi-objective optimization, and external knowledge systems into model designs. Innovate methods to interoperate with LLMs, LRMs, LMMs, and transformer-based architectures, ensuring seamless integration and collaboration within the AI ecosystem using AI Multi-Agentic Frameworks. Model Development : Build and refine Bayesian or Markovian Graph chains to incorporate uncertainty estimation, adaptive decision-making, and probabilistic reasoning. Develop foundational models by merging Bayesian techniques with Classical ML with L[L/M/R]Ms