A sushi chain is leveraging Anthropic's Claude Sonnet 4.6 to manage customer interactions via Instagram DMs, handling approximately 90% of orders through this channel. The system's affordability is primarily due to an effective caching mechanism, which serves the extensive menu and operational context from cache for about 97% of messages, significantly reducing processing costs. This agent assists customers with order selection, provides information on ingredients and allergens, suggests upsells, and integrates confirmed orders into the kitchen and CRM systems, while complex requests like photos or voice notes are routed to human staff. AI
IMPACT Demonstrates how caching strategies can make LLM-powered customer service affordable for businesses with high interaction volumes.
RANK_REASON This is a case study of a specific business implementing an LLM for customer service, not a direct release from a frontier lab.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
