Social Media Bot
Scenario:
- KLM handling 15,000 customer service cases per week from social media
- When an agent receives a customer question, the AI proposes an answer
- Agent decides whether the proposed answer is accurate, adjusts it if necessary and sends the response
Implementation:
- Integrated into KLM’s Customer Relationship Management tool (Salesforce)
- Trained on 60,000+ questions and answers. The system learns based on what the agent does.
- Uses DigitalGenius
Google assistant
- With Google Assistant's KLM bot users receive tips on how to pack bags for flight based on a destination.
- Uses Artificial Intelligence through Blue Bot service to speak to users in a conversational manner.
- Answers frequently asked questions and provides information about airline requirements.
Facebook Messenger Chatbot
- KLM named their chatbot BB so that it would have a name like any other person
- BB's personality is female, helpful, friendly, and professional
- The chatbot books flights, sends confirmations, and handles payment
- By connecting BB with KLM's CRM system, a human agent can take over if BB is unable to answer
- Impact:
- 60,000 questions answered
- Receives around 100,000 mentions per week
- 15,000 conversations per week
- The app was built using Google Dialogflow
Tags: Travel, Facebook Messenger, Twitter, Google Assistant