Senior Machine Learning Scientist, Agentic AI & Genai Systems (growth Marketing)

Expedia Expedia · Hospitality · Seattle, WA

Senior Machine Learning Scientist role focused on building end-to-end agentic AI and GenAI systems for growth marketing at Expedia. The role involves architecting, building, and shipping multimodal LLM systems, RAG pipelines, multi-agent ecosystems, and integrating them with user interfaces and observability features. It also includes training/fine-tuning LLMs, developing memory architectures, and deploying large-scale behavioral embedding systems.

What you'd actually do

  1. Architect, build, and ship enterprise-scale GenAI, RAG, and multi-agent systems end-to-end, including frontend, backend, and user interfaces.
  2. Design hierarchical multi-agent ecosystems with Interactive Generative UIs, dashboards, and safety/observability features (e.g., UI is generated on-the-fly by an agent in response to what the user need).
  3. Develop memory architectures: short-term contextual memory, long-term episodic memory, knowledge graph augmentation, and adaptive retrieval systems.
  4. Lead hands-on implementation of RAG pipelines, vector memory systems, and agent orchestration frameworks (LangChain, LangSmith, AutoGen, OpenAI/Claude Agents SDK, etc) and advanced evaluation platforms leveraging LLM-as-a-Judge, synthetic data generation, and agent trajectory assessment.
  5. Train, fine-tune, adapt (LoRA/QLoRA/adapters), and distill LLMs, including RLHF/DPO, for production-ready chatbots and GenAI products.

Skills

Required

  • 10+ years in software engineering, ML, and AI systems, with production GenAI deployments.
  • Deep expertise in LLM training, adaptation, distillation, RLHF/DPO, and RAG systems.
  • Solid foundation in NLP and experience with multimodal AI systems (vision-language models)
  • Proven experience building and operating multi-agent AI platforms with observability and safety frameworks (self-hosted orchestration using frameworks such as LangGraph integrated with LLM APIs such as Claude).
  • Strong background in distributed GPU training and inference, cloud infrastructure (AWS/Azure), container orchestration, and ML tooling.
  • Demonstrated ability to lead end-to-end AI product development and collaborate with product and design teams to ship user-facing features.
  • Excellent communication skills, able to present complex architecture and product concepts to executives.

Nice to have

  • PhD in Computer Science, Machine Learning, or a related field.
  • Recognized industry presence through publications, patents, talks, or open-source contributions in LLMs, RAG, or agentic systems.
  • Experience integrating multimodal LLM systems (vision, audio, music, structured data).
  • Leadership in GenAI safety, evaluation, testing, and monitoring.
  • Strong cross-disciplinary fluency in modeling, infrastructure, product, and design.

What the JD emphasized

  • building end-to-end systems
  • backend architecture to user-facing interfaces
  • agentic UX/UI, observability, and product integration
  • Interactive Generative UIs, dashboards, and safety/observability features
  • generated on-the-fly
  • multimodal pipelines integrating vision, audio, text, and structured data
  • scalability and latency
  • production GenAI deployments
  • multi-agent AI platforms with observability and safety frameworks

Other signals

  • building end-to-end systems
  • agentic UX/UI, observability, and product integration
  • multimodal LLMs, GenAI, and agentic architectures
  • production AI
  • multi-agent AI platforms