Retail
Illustrative example
An assistant that resolves six in ten customer support requests
An AI assistant connected to the order system and knowledge base, handing over to the team on sensitive cases.
- Client
- National furniture retailer, 40 stores
- Sector
- Retail
- Project type
- AI implementation
- Duration
- 9 weeks
Results
−58%
average response time
61%
requests resolved without intervention
9 sem.
from diagnosis to production
The challenge
The support team handled thousands of requests a month, mostly about order status, deliveries and returns. Response times grew during campaigns and the team had no room for the cases that truly needed attention.
The approach
Diagnosis: analysis of six months of requests to identify the recurring topics.
Plan: three request types chosen for the pilot and the metric defined — average response time.
Implementation: a four-week pilot on one channel, with the team validating answers, before rolling out to the rest.
The solution
An assistant that checks order status in real time, answers from company policy and opens the returns process. When it detects dissatisfaction or a case outside the rules, it hands the conversation to a person with the summary already written.
“The team stopped answering "where is my order" and started looking after the customers who really need us.”