The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Building a Recommendation Agent for The North Face with Andrew Guldman - TWiML Talk #239
Mar 14, 2019
Andrew Guldman, VP of Product Engineering and R&D at Fluid, shares insights into creating AI-driven user experiences for online retail. He discusses the innovative Fluid XPS, developed for The North Face to simplify outdoor gear selection for casual shoppers. Guldman delves into using advanced algorithms and graph databases, the challenges of keeping pace with evolving AI, and the balance between programming and flexibility. His anecdotes highlight the importance of emotional connections in product recommendations, making tech more human-centered.
AI Snips
Chapters
Transcript
Episode notes
XPS origins
- The North Face initially targeted expert climbers, runners, and skiers.
- They wanted XPS to help casual users, like first-time backpackers, choose the right gear.
XPS Functionality
- XPS emulates a sales associate by gathering information through conversation.
- It uses a graph database and heat-sink algorithm to recommend products based on user facts.
AI in XPS
- XPS was initially intended to use Watson but its single-answer API wasn't suitable.
- They shifted to a content-based recommendation engine with NLP, aiming for collaborative filtering later.