ALF

ALF is an AI agent that handles simple and repetitive inquires for you. Depending on the question and context, it can autonomously suggest features that may be useful to users.

• ALF handles simple and repetitive inquiries, freeing agents from the task of verifying and managing them. This reduces the agents' workload on straightforward questions, giving them more time to focus on complex issues.

• ALF can utilize the FAQ commands, FAQs, and RAGs described below to proactively manage chats.

Before introducing ALF, we'll cover some important terminology to help you better understand and utilize this AI agent.

• Command: Commands are “Instructions” that perform a specific action.

  • ALF can automatically answer by selecting the appropriate command from the command list to handle the customer's inquiry

  • Customers can manually access commands by typing “/” (forward slash) followed by command.

• FAQ : You can provide the most frequently asked questions and their answers, and ALF will respond to those for you. (→ View our FAQ Guide)

• RAG (Search Augmented Generation) Coming Soon : AI will scan pre-registered manuals to find answers to customers' questions and generate a response accordingly. This capability will also be integrated into ALF in the future.

1. Activate Command: This feature allows ALF to find suitable commands and provide responses accordingly.

2. Activate FAQ: This feature enables ALF to utilize pre-set FAQ responses to answer customers' inquiries.

3. Describe Persona: This feature aims to align ALF with the expected persona, enabling it to engage customers with more appropriate responses tailored to the designated personality. It helps prevent common hallucination* phenomena in AI models.

   (Hallucination*: The phenomenon where an AI model generates inaccurate or fabricated information that is not true.)

   Example:

   • For a Fashion brand: “ALF serves as the customer service representative for Fashion Brand A. It can assist with inquiries related to shipping, refunds, exchanges and cancellations.”

   • For a hospital - “ALF is in charge of hospital B’s customer support. It can assist with managing bookings and visit dates.”

   • For an education facility: “ALF is the academy representative of school C. It can assist with term start dates, completion requirements and discounts.”

   

• You must be using a paid plan + support but as ALF can only be activated at the end of a support bot.

• You can activate ALF at the end of each support bot step.

  • Example

    • You can activate ALF for simple inquiries like Shipping and Refunds

    • but if it’s a stage where consultation, like an introductory conversation, is needed, you can simply turn ALF off.

Will there be additional charges?

• ALF will be offered as a free feature temporarily, and will become a subscription-based feature in the future. We’ll keep you informed of any updates regarding this transition.

What sets ALF apart from other AI solutions?

• While AI like ChatGPT answers questions based on general knowledge, ALF is a dedicated AI agent for Channel Talk, capable of providing tailored responses to customers through commands and FAQs. Channel Talk is the only solution that applies AI to all stages of customer consultation.

What is the difference between FAQ and RAG?

• FAQs delivers responses exactly as you have written in the [Answer] in a FAQ. RAG allows ALF to reconstruct their response based on the manual given.

Can ALF be turned on or off all at all stages at once?

• Turning ALF on or off all at once is not supported.