Machine Learning Engineer, Assistance Engineering
Machine Learning Engineer on Airbnb's Assistance Engineering team building Agentic AI for customer service, including the Chat AI and Voice AI assistants. Day to day: developing LLM-based systems (fine-tuning, RLHF, RAG, multi-agent orchestration) and shipping production ML services with product and design partners.
- Compensation
- $162K–$186K / yr
- Location
- Remote-USA
- Work setup
- Remote
- Seniority
- Not specified
- Employment
- Not specified
- Experience
- Not specified
- Category
- Ml / AI
- Posted
Skills
Must have
- PhD or Master's degree with 3+ years of experience in Computer Science, Machine Learning, AI, or related technical field — or equivalent industry experience
- Hands-on expertise in LLM: pretraining, fine-tuning (SFT, RLHF, GRPO), prompt engineering, RAG architectures, and LLM evaluation frameworks
- Experience building Agentic AI systems — multi-agent orchestration, tool-use, planning, memory, autonomous reasoning pipelines (e.g., ReAct, LangGraph, AutoGen)
- Experience shipping production-grade ML/AI systems at scale, with deep understanding of ML infrastructure, model serving, and MLOps best practices
- Excellent communication skills with ability to collaborate across Engineering, Product, and Design
- Must live in a US state where Airbnb, Inc. has a registered entity
Full description
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.
The Community You Will Join:
Machine Learning and Artificial Intelligence are at the heart of the Airbnb product. From Trust to Payments, and from Customer Service to Marketing we rely on ML to ensure that guests and hosts have the best possible experience with Airbnb.
The Core ML team in Assistance Engineering is the team responsible for adopting the Agentic AI technologies to enable an intelligent, scalable and exceptional customer service experience. We are responsible for developing the Chat AI assistant, Voice AI Assistant and more! The team is constantly exploring the SOTA Agentic architecture, develops and enhances various AI models, ML services and leverages tools including SFT, Reinforcement learning, Distillation, RAG/Search, LLM evaluation and testing automation, feedback-based learning and guardrail for a wide range of applications in Airbnb.
The Difference You Will Make:
We believe our current customer experiences in these domains are only scratching the surface of the innovations that are possible, and that science is at the heart of delivering a step-function change for our Guest and and Host on Airbnb.
You will build and leverage cutting edge AI technologies to transform Airbnb’s customer service by delivering personalized, easy-to-use and proactive customer service experience.
Many of the initiatives you’ll tackle are in their early conceptual stages. You will have the opportunity to shape these ideas from inception to production, turning visionary concepts into impactful realities.
A Typical Day:
- Champion the development of novel ML systems, product integrations, and performance optimizations to solve real-world problems
- Work cross-functionally with product, design, and other engineering counterparts to design and build efficient AI solutions for Airbnb CS products
- Learn and share the latest AI/ML technologies with the team.
Your Expertise:
- PhD or Master's degree w/ 3+ YOE in Computer Science, Machine Learning, Artificial Intelligence, or a related technical field — or equivalent industry experience
- Hands-on expertise in LLM, including pretraining, fine-tuning (SFT, RLHF, GRPO), prompt engineering, RAG architectures, and LLM evaluation frameworks
- Experience building Agentic AI systems — including multi-agent orchestration, tool-use, planning, memory, and autonomous reasoning pipelines (e.g., ReAct, LangGraph, AutoGen, or similar)