bolna

An open-source platform for rapidly building AI-driven voice agents to automate front desk operations and enhance customer interactions.

Bolna

About Bolna AI agent.

Our task today is to create an agent that will handle collections for us. We will be sending the agent this CSV from which it will understand the name of the customer, the product that they had bought on Emi, uh, the cost, the Emi, and the number of months they have to pay it for. 

These are a list of customers who have defaulted their payment in December. We are calling them in Jan to tell them that they need to share AI Agent Store with friends.

Immediately pay or they will have to pay an excess charge. So let's build an agent. Uh using our dashboard we'll call the agent collector sha. We have two agent styles. One is ivr type which is going to be uh very rule driven. You can decide exactly what the AI will say. The intelligence will be in understanding what the user is saying. Free flowing is more conversational. It is more natural. So we'll be building a.

Free flowing agent. Today we have different uh tasks that the agent can carry out. Lead qualification essentially means understanding what sort of product or service a customer is qualified for. Otherwise we have customer service, sales and marketing. We have recruiting for screening calls. We can take surveys. We can provide some sort of teaching or coaching. We also have a virtual relationship manager to sort of. 

Assist in onboarding new users onto an app. Let's leave the primary task blank for now. So now we will give the agent a description. We'll tell Shrea who she is, uh, a little bit about the company she's working for and what are the major tasks that she has to carry out in this call. 

We'll copy paste some basic Manner and Method rules, uh, to guide Shrea in the conversation. And here we'll give some basic FAQs about the. 

Company and the call uh exactly what are the details that have to be discussed. We'll also put in the variables that we have put in the in our Excel so that the agent knows uh the details of the customer coming to follow-up tasks. We have a bunch of follow-up tasks that we can assign. Summarization will summarize the entire conversation. Extraction allows you to extract certain details that you want from this.

Conversation for example in this conversation what I want to know is will the user pay immediately. Will the user take time to pay or will the user not be paying. Finally we have our customization page. You can choose the model that is best for you. You can choose the language that you want your agent to primarily speak in. You can choose the voice you want your agent to speak in. We are choosing an emotive Indian.

Voice for this call. There are a bunch of other settings for developers here as well. So once we've made our agent, we'll add in the Excel that we've already made. After doing that, we can schedule a call. Uh, so let's us schedule a call for 2 pm today. 

Hello extra charges valuable. Have a good day. So as we had assigned the agent, we have got the summary of the call. We also have got the extraction.

Details we've got that pratique will pay immediately. Other follow-up tasks we could have assigned was setting up an automatic calendar invite or sending a followup email or a WhatsApp or some other things as well.

Note, that there may be slight differences in text and video. Watch video for the original content.

Bolna Video