Cadenflow

Cadenflow successfully took on the project of an AI Customer Support System for an ecommerce company that sells healthcare products, where they aimed to reduce operation costs and save time. The system was designed to support the company's increasing customer orders that were proving to be unsustainable for their small internal team. It involved an AI-powered service that enabled automatic email reading, customer and order identification, request classification, auto-drafted replies, human approval workflow, safety guardrails and staged rollout plan, all while ensuring GDPR compliant handling of customer data.

From the inception, Cadenflow engaged in a thorough study of the company's support inbox. They sifted through approximately 400 previous support conversations, categorizing each ticket based on intent. This provided the client valuable insight into what customers typically contact them about. The AI's auto-drafted response feature was an intriguing aspect for the ecommerce business, as it could draft replies in multiple languages such as English, German, Spanish, Italian, and French, in their brand voice using information from their return policy, product FAQs, and a specific tone-of-voice guide.

The project has not only benefited the ecommerce company through reduced operational costs, but also improved their support system by automating the triage of support tickets and streamlining their workflow. The staged rollout plan designed by Cadenflow ensured that new features were only introduced once they reached a defined accuracy threshold, maintaining a balance between efficiency and quality. All sensitive cases were flagged and routed to human support for review, ensuring no automated errors. The client highly valued the weekly analysis of system performance against real tickets provided by Cadenflow, which was backed by documented findings and resultant fixes, ultimately fostering a trust-based relationship with the service provider.