How Do We Test our AI Chat and Voice Bot?
Valliammai Senthilnathan and Eugenia Avramenko
Electrolux
We work in the Service and Repair industry of Electrolux. So we have AI chat and voice Bot which the consumer can interact with when they have any questions related to service and repair. AI chat bot saves time and direct the call to the human agent with the proper skills and would also have collected basic information of the consumer for the agent to refer and get back quickly.
First thing here is to understand what the value is the AI bot is providing and then we need to ensure that we are covering that as part of testing. Being we wanted to assure a great consumer experience its very important to consider nonfunctional aspects of the BOT along with functional aspects. The following functionalities of chatbot will be tested.
1. Response time
2. Answering/ response accuracy
3. Error management
4. Personality
5. Navigation
6. Intelligence
7. Understanding
8. Multi/cross platform & device compatibility
We use botium tool to test our AI chatbots. We would like to explain a detailed test process and challenges with respect to AI chatbot testing.
This is part of AI Deep Dive. Check out our other talks.
About Me!
Valliammai Senthilnathan is a highly skilled and results-driven professional with 16+ years of experience in quality assurance and validation—combined with a dedication to product quality excellence. Her background includes outstanding experience in leading the testing team on agile train and providing solutions on more strategical level to drive continuous quality improvements and ensure compliance with regulations.
Eugenia Avramenko;s background includes outstanding experience in leading the testing team on agile train and providing solutions on more strategical level to drive continuous quality improvements and ensure compliance with regulations.