1. How is generative AI changing the use of chatbots?

It is improving substantially because we are now able to understand much better all the nuances of each question. The novelty really lies in the fact that the answer no longer has to be the same every time for the same type of question. The response will adapt to the user's needs, even to the tone, and it can link references that may vary depending on how the question was phrased. As a result, we are going to see chatbots that increasingly have a higher level of expressiveness.

2. What advantages can a small business expect when integrating a chatbot into their operations?

It always has to do with 24/7 availability, right? Thinking about those times of the day when you don’t have human support available, a chatbot, for example, can serve as a first triage for the issues you receive. It can also serve as a pulse check, identifying what people are asking about the most, mapping the chatbot with analytics to see when most questions are asked and where there are more doubts. This can also allow you to refine the information you provide as a service. Another thing that a chatbot can help with, which might not be so obvious, is handling multiple languages if you’re considering international expansion. For example, having a chatbot that helps with automatic translation of content for users who don’t speak the same language as you can be very practical, especially for breaking down that initial barrier.

3. What are the first steps a SME should take when integrating a chatbot?

When integrating a chatbot, several factors need to be considered. First of all, you need to be clear about the type of task you want it to handle. The chatbot should be aimed at, say, providing support, helping with a purchase process, or simply resolving very specific issues. Identifying this is crucial because we will then need to work on the content. This does not exempt us from our role as humans. Therefore, it’s important to have the task clearly identified. Another common mistake when setting up a chatbot is what comes afterward, which is the validation system. We will not only need response content but also to train it later with different people interacting with the chatbot. So, it’s a good idea for the company itself, the small business, to prepare a set of questions that can later validate the chatbot's proper functioning

4. What low-cost tools can be used to create a chatbot?

Dialogflow CX is a tool by Google, which even has a fairly extensive free tier. The great thing about Dialogflow CX is that it already incorporates the ability to use generative AI. It’s what we call a No-code or Low-code tool. Essentially, it’s built around creating the chatbot using conversation boxes. We create these conversation flows very easily. Then, for example, in moments where the chatbot doesn’t know how to respond or where we want it to be a bit more expressive, we can add these boxes that will call a generative AI model to provide a more tailored response for the user.