TextYess is a fast-growing, WhatsApp-focused AI shopping assistant that enables eCommerce platforms to turn conversations into orders. As TextYess and the number of interactions on their LLM-based shopping assistant grew, flagging suboptimal LLM responses became a challenge.

What TextYess realized they needed:
‍
• Scalable visibility into LLM user interactions
‍
• Flagging each suboptimal response
‍
• Pinpointing and understanding successful user interactions to replicate high-converting experiences

With Nebuly, TextYess is able to scale its user base and business while having a consistent overview of the quality of its shopping assistant user interactions.

Nebuly's user feedback feature automatically flags each negative and positive (not neutral) LLM interaction for TextYess. The insights allowed them to end manual output reading, pinpoint areas of improvement and continuously improve their shopping assistant's user experience.