Free form slot capture examples

AI-generated summary

Genesys Cloud's Architect platform integrates advanced slot management capabilities with AI-powered functionality for Virtual Agent interactions. The system employs four custom slot types - list, dynamic list, regular expression, and AI-powered slots - enabling sophisticated data capture and natural conversational flows. Multi-turn conversations allow flexible information collection, while enhanced features support name spelling recognition, country-specific address formatting, and email validation. The platform excels in handling complex data corrections through natural dialogue, allowing users to modify individual components of names, addresses, and email addresses while maintaining contextual awareness. Notable capabilities include multi-part address collection, spelled-out name processing, and various email format handling. The system demonstrates particular strength in incremental information gathering and correction, supporting both complete and partial data updates. Key limitations include country-specific address format variations, inconsistent text casing, custom domain email recognition challenges, and accuracy issues with special characters and spelled-out names. Despite these constraints, the platform effectively manages data validation, correction flows, and confirmation processes, ensuring accurate information capture across different interaction types.

Architect supports four custom slot types: list slot type, dynamic list slot type, regular expression slot type, and AI-powered slot type for Virtual Agent. If you enable Virtual Agent, you can create generative, AI-powered slot types when you create a slot from the Natural Language Understanding menu.

AI-powered slots help Virtual Agent administrators and flow authors to capture and collect multiple slots in any order, capture slots over multiple turns, and make it easier for the bot to understand when customers perform tasks that are more complex for the bot, such as spell out their name. When you enable Virtual Agent, you can use numeric, alphanumeric, or free-form AI-powered slots and slot types to improve conversational bot behavior. These options allow you to configure your bot so that conversations between the bot and participant more closely align with human conversations.

This article describes limitations and examples of AI-powered free-form slots. For more information about numeric and alphanumeric AI-powered slots, see Use AI Powered slot types in a bot flow.

Free-form slot limitations

Addresses
- Address format per specific country standards. The bot participant must rely on the description that you provide to ensure the proper format.
- Casing: The correct casing is typically correct; however, the extracted entity may occasionally return as all lowercase or all uppercase.
Emails
- Incorrect identification of custom domain names across multi-turn conversations. More common, accurate returns occur when the emails are provided in a single turn.
- Cases where ASR transcriptions fail to convert dashes, dots, and underscore.
Names
- Missing or rearranged characters when long names are spelled out.

Free-form slot basic examples

Name	Description	Conversation
Name	Full name of a customer	Bot: Please provide a value for name. Customer: Grace Diley Bot: Thanks for providing name: Grace Diley
Name	Full name of a customer	Bot: Please provide a value for name. Customer: It’s Little. Bot: I got Little so far, please continue. Customer: Last name’s Simz. Bot: Thanks for providing name: Little Simz.
Address	Delivery address of a customer	Bot: Please provide a value for address. Customer: One two three Main Street, Galway. Bot: Thanks for providing address: 123 Main Street, Galway.
Address	Delivery address of a customer	Bot: Please provide a value for address. Customer: Two zero two Parkview terrace. Bot: Thanks for providing address: 202 Parkview Terrace.
Email	Email address of a customer	Bot: Please provide a value for email. Customer: It’s JDV 123 at gmail dot com. Bot: Thanks for providing email: jdv123@gmail.com.
Email	Email address of a customer	Bot: Please provide a value for email. Customer: Catlover at hotmail dot com. Bot: Thanks for providing email: catlover@hotmail.com.

Free-form slot complex examples

Name	Description	Conversation
Name	Full name of a customer that includes a first name and a last name	Bot: Please provide a value for name. Customer: It’s a for apple, n for not, a for apple and then my last name is branden. Bot: Thanks for providing name: Ana Branden.
Name		Bot: Please provide a value for name. Customer: It’s a for Apple, m for Mango, b for Ball, e for Egg, r for Run. Bot: I got Amber so far, please continue. Customer: Dee’s my last name. Bot: Thanks for providing name: Amber Dee.
Address	Delivery address of a customer that includes a house number and a PIN code	Bot: Please provide a value for address. Customer: It’s 77 Woodquay Court, Galway. Bot: I got 77 Woodquay Court, Galway so far, please continue. Customer: I think the eircode is T12H6EK. Bot: Thanks for providing address: 77 Woodquay Court, Galway T12H6EK.
Address		Bot: Please provide a value for address. Customer: House number is 66. Bot: I got 66 so far, please continue. Customer: It’s near Clifden in Co. Galway, Ireland. Bot: I got 66, Clifden, Co. Galway, Ireland so far, please continue. Customer: Abbey Road, and the PIN code is T12H6EK. Bot: Thanks for providing address: 66 Abbey Road, Clifden, Co. Galway, Ireland T12H6EK.
Email	Email address of a customer that includes a username and a PIN code	Bot: Please provide a value for email. Customer: It’s q for queen, w for win, e for egg, r for Rome, t for tiger and y for yacht at gmail.com. Bot: Thanks for providing email: qwerty@gmail.com.
Email		Bot: Please provide a value for email. Customer: P j underscore smith. Bot: I got pj_smith so far, please continue. Customer: At yahoo.com. Bot: Thanks for providing email: pj_smith@yahoo.com.

Free-form slot correction examples

Name	Description	Conversation
Name	Full name of a customer that includes a first name and a last name	Bot: Please provide a value for name. Customer: First name’s Tirana and my last name is Groj. Bot: Thanks for providing name: Tirana Groj. Customer: Ah the first name is Dirana actually. Bot: Thanks for providing name: Dirana Groj.
Name		Bot: Please provide a value for name. Customer: Manan. Bot: I got Manan so far, please continue. Customer: Calra. Bot: I got Manan Calra so far, please continue. Customer: Ah sorry, the last name starts with a K. Bot: Thanks for providing name: Manan Kalra.
Address	Delivery address of a customer that includes a house number and a PIN code	Bot: Please provide a value for address. Customer: It’s 77 Woodquay Court, Galway. Bot: Thanks for providing address: 77 Woodquay Court, Galway. Customer: Goodquay court, not wood. Bot: Thanks for providing address: 77 Goodquay Court, Galway. Customer: Ah sorry, the house number is 87. Bot: Thanks for providing address: 87 Goodquay Court, Galway.
Address		Bot: Please provide a value for address. Customer: Eight seven six Merchants Road, Galway. Bot: I got 876 Merchaps Road, Galway so far, please continue. Customer: It’s Merchants, not Merchaps. Bot: I got 876 Merchants Road, Galway so far, please continue. Customer: T12H6EK, that’s all. Bot: Thanks for providing address: 876 Merchants Road, Galway T12H6EK.
Email	Email address of a customer that includes a username and a PIN code	Bot: Please provide a value for email. Customer: My email is qwerty125@rediffmail.com. Bot: Thanks for providing email: qwerty125@rediffmail.com. Customer: Ah it’s hotmail.com sorry. Bot: Thanks for providing email: qwerty125@hotmail.com. Customer: Oops, it’s supposed to be qwerty129. Bot: Thanks for providing email: qwerty129@hotmail.com.
Email		Bot: Please provide a value for email. Customer: qwerty728. Bot: I got qwerty728 so far, please continue. Customer: Gmail.com. Bot: Thanks for providing email: qwerty728@gmail.com. Customer: Not gmail, it should be yahoo. Bot: Thanks for providing email: qwerty728@yahoo.com.

[NEXT] Was this article helpful?

Get user feedback about articles.