Bots failed on complex returns, forcing calls. AI agents now use live supply chain data to resolve 60% of WhatsApp inquiries.
A leading Latin American retailer serves 36 million customers across seven countries, managing diverse business lines from supermarkets and home improvement to financial services.
Rising digital sales drove a spike in inquiries, but legacy chatbots could not distinguish complex return policies for refrigerators versus t-shirts....
“Returning a refrigerator or a TV is different from returning a T-shirt, or a carton of milk. There are a lot of processes, definitions, and policies around what you can return, when, and how depending on the product. And solutions are different in every country. Training a human being to really understand what to do in every single situation is extremely difficult.”
Multinational department store and e-commerce retailer in Latin America.
AI-powered CRM platform for sales, service, marketing, and commerce teams.
Related implementations across industries and use cases
IT reporting took 10 mins while shoppers waited. AI now captures tickets via voice or text, cutting resolution time from 2 hours to 30 mins.
Generic queries clogged phone lines. AI now resolves 60% of app inquiries, freeing advisors to solve complex customer issues.
Classifying 100 comments took an hour, forcing sampling. Llama 3.3 now sorts every review by issue type to prioritize fixes.
Generic queries clogged phone lines. AI now resolves 60% of app inquiries, freeing advisors to solve complex customer issues.
Rigid menus trapped users in queues. An AI agent now resolves inquiries on WhatsApp instantly, boosting response accuracy 109%.
Store managers lost applicants to a 10-day hiring cycle. Now, an AI agent screens and schedules via text, saving thousands of hours.
Shoppers built lists item-by-item. Now, they build full baskets in one prompt, while staff automate internal workflows.
Inboxes overflowed with tracking and sizing queries. An AI now locates orders and asks player height to size gloves instantly.
Specs scattered in PDFs meant fixes took weeks. Now, autonomous agents access unified data to resolve over 25% of cases instantly.