Plush is looking to improve its core search functionality, build personalized discovery systems, and enable a real-time, chat-based personal stylist that understands user intent and returns highly curated, shoppable results to reshape how people discover fashion online.
Requirements
- 3+ years of experience working on machine learning, search, and recommendation systems.
- Comfortable designing both quick prototypes and scalable infrastructure.
- Strong product intuition and care about what users experience.
- Built LLM agents or RAG pipelines in production.
Responsibilities
- Improving our hybrid search systems: Combine keyword and semantic search to surface the most relevant fashion results—balancing relevance, diversity, personalization, and business priorities (e.g., store boosting).
- Building ML systems that learn user preferences: Tailor search results—surface brands they love, styles that match their event or mood, and pieces that complete an outfit based on their browsing and purchase history.
- Exploring social discovery features: Power community-driven recommendations and enhance product discovery (e.g., “people with similar taste loved this” or “you and your friend share a love for minimalist tailoring”).
- Establishing evaluation pipelines: Measure and improve search and LLM performance—tracking relevance, coverage, latency, and output quality.
Other
- High ownership, clear communication, and thrive in fast-moving environments.
- Genuine interest in fashion and how people discover what to wear.
- Authorized to work in the U.S. or Canada and can work within U.S. time zones (remote-friendly within that constraint).
