2 months ago
Responsibilities
- Define architecture and contracts for how models move from development to production, including feature stores, model schemas, inference consistency, and multi-version support.
- Lead development of a unified serving stack that provides a standardized path from model training to production.
- Architect backfill and evaluation infrastructure for simulating production inference over historical data.
- Establish clear, enforced contracts between modeling and serving teams.
- Review and evolve ML serving architecture and make tradeoff decisions on feature pipelines, model composition, and API interfaces.
- Write and review code for feature engineering jobs, feature store configurations, and serving service endpoints.
- Partner with Data Science, MLE, MLI, and Pricing & Availability backend teams on artifact handoffs and integration contracts.
- Drive milestone planning across the Host Pricing & Settings organization and mentor engineers through design reviews and hands-on pairing.
Requirements
- 12+ years of backend or platform engineering experience with substantial experience building production ML systems or data-intensive infrastructure.
- Strong programming skills in Java, Kotlin, Scala, and/or Python.
- Deep understanding of feature stores, training/serving consistency, model versioning, and online/offline inference pipelines.
- Experience with high-scale batch and real-time data pipelines, including point-in-time correctness for backfills.
- Expertise designing APIs, efficient data contracts, and multi-tenant serving infrastructure for large-scale applications.
- Proven ability to lead cross-team technical initiatives spanning ML and platform engineering.
- Preferred production experience with Chronon, Tecton, Feast, or equivalent feature store technology.
- Preferred experience with model schema management, multi-version support, model composition frameworks, domain contract design, and improving ML model evaluation and production deployment velocity.
Benefits
- US remote eligible
- Occasional work at an Airbnb office or attendance at offsites may be required
- Must live in a state where Airbnb, Inc. has a registered entity
Tech Stack
Categories
About Airbnb
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, experiences and services that make it possible for guests to connect with communities in a more authentic way.