Frequently asked questions about conversational agents and chatbots
24 questions
Conversational AI → What is a conversational agent?
It’s an AI system that uses conversation as the interface to look up information, use tools or carry out actions.
What’s the difference between a chatbot and a conversational agent?
What matters is whether it acts, not what it’s called. A scripted chatbot sticks to prepared answers. A conversational agent, or an AI chatbot connected to the company’s systems, looks up information and completes processes.
Can a chatbot query a CRM?
Yes. It can retrieve authorized information from a CRM and use it within the conversation.
Can a chatbot carry out actions?
Yes. It can create records, start workflows, change appointments or carry out other actions within the defined permissions.
Can a conversational agent work on WhatsApp?
Yes, as long as there’s a proper integration with the channel.
Can it work on a website?
Yes. It can also be integrated into applications, internal portals or other channels.
Can it remember the context of a conversation?
Yes. It can keep relevant information within the session or use additional mechanisms when the case requires it.
What happens if the chatbot doesn’t know the answer?
It can ask for clarification, say it doesn’t have enough information or hand the conversation to a person.
Can it transfer a conversation to a human?
Yes. It can do so while keeping the context, the data collected and the status of the process.
Can it use internal documentation to answer?
Yes. It can connect to RAG systems, knowledge bases or other authorized sources.
How do you stop a chatbot from making up answers?
By limiting sources and using information retrieval, checks and rules. For when it has no answer there is a planned alternative, such as handing the conversation to a person.
Can a conversational agent replace a support team?
Not necessarily. It’s usually more useful for automating repetitive questions and leaving complex cases to the human team.
Can a chatbot identify a user?
Yes, if the channel or system has an authentication or identification mechanism.
Can a chatbot show personalized information?
Yes. Once the user is identified, it can look up authorized information linked to their account.
Can a conversational agent work with voice?
Yes, if it’s integrated with speech recognition and speech synthesis.
Can a chatbot handle forms through conversation?
Yes. It can collect the data one field at a time and structure it automatically.
Can a conversational agent understand ambiguous messages?
It can try, and ask for clarification when the intent isn’t clear enough.
Can it detect changes of intent during a conversation?
Yes. It can update the flow when the user changes their goal.
Can a chatbot handle several languages?
Yes, as long as the models and sources are prepared for it.
Can a conversational agent pick up a previous conversation?
Yes, if the design includes persistent memory and it’s allowed from a privacy and product standpoint.
Can a chatbot send files or links?
Yes, if the channel used supports that type of content.
Can a chatbot check availability in real time?
Yes. It can do so by connecting to calendars, inventory or other systems.
Can a chatbot take a payment or start one?
It can start a payment flow or generate a link, although financial transactions should go through specialized secure providers.
Can a conversational agent qualify leads?
Yes. It can collect information, apply criteria and log the opportunity in a CRM.
Tell us about the process you want to improve.
You do not need to know which model to use, which architecture you need or whether the problem calls for an AI agent. Tell us how the process works today, who is involved and which tools you use.
We’ll reply within one business day with a first read of your case.