A Beginner's Guide to AI The AI Stylist for Men: AI Can Dress You Better Than You Do - says Zoher Karu
Feb 27, 2026
Zoher Karu, head of AI at Taelor who blends algorithms with human stylists, explains using AI to simplify men’s wardrobes. He discusses style profiles, LLM-driven outfit logic with human oversight, handling nonstandard sizing and rich garment metadata. Conversations cover vector matching, reducing decision fatigue, and scaling personalized menswear through feedback loops.
AI Snips
Chapters
Transcript
Episode notes
LLMs Provide Styling Intuition But Need Constraints
- Taelor uses large language models to reason about which items go together, then refines outputs with proprietary styling rules.
- LLM suggestions are constrained by inventory and then adjusted by human stylists for the target demographic.
Fashion Is A Multi‑Dimensional Matching Problem
- Fashion is a complex matching problem because you must model the customer, context, and garment properties simultaneously.
- Zoher calls it like "a dating website gone on steroids" where taste, occasion, fit and fabric all matter.
Measure Clothes Not Just Size Labels
- Measure garments and capture detailed metadata instead of relying only on brand sizes to improve fit predictions.
- Taelor physically measures items because a "small" varies across brands and customer feedback refines the model.