You've been assigned a new customer, Contoso LLC, which sells bicycles and bicycle equipment to its customers. Contoso currently processes new product orders and queries through human operators. They are looking for an automated solution that allows Contoso to seamlessly scale up to handle large call volumes while maintaining zero wait times and freeing up staff to manage other tasks.
Your team recently presented a potential architecture (below) that Contoso LLC approved.
- Skype Client
User initiates call - Bot Connector + Skype Calling Channel Routes calls from Skype to the bot
- Azure App Services
Hosts the bot application, which manages logic and API calls - Cosmos DB
Stores bot state and event logs - Speech Service Processes speech-to-text
- LUIS (Language Understanding Intelligent Service)
Extracts intent and entities from text - Azure Search
Indexes the product catalog for product-query matching - Azure SQL
Stores product and order data - Azure Storage
Stores bot audio data for debugging
Your team is preparing a proof of concept, and your group has been tasked to suggest answers to the following questions:
- What might your main intents in LUIS be?
- Are there any additional Cognitive Services that you think could be used to help bring value to the business?
- What are some potential ideas for how Custom Vision could be used to bring value?
Document your suggestions on the whiteboards/flipcharts provided. You will present your results.