Launching Autonomous Collections 2.0 | HighRadius

Speaker 1 [00:05]
Autonomous collections. We start off at that bottom piece of domain functionality, which is your collection strategies, payment commitments, dispute management, and then our data driven AI is built on top of that, where we have human propensity, customer segmentation, risk engaging and customer intent. Then from there, taking that next step with that integration around email and consent systems, automate correspondence, portal integrations, notes and actions, and then taking that experience and bringing that to life for a user experience with free to the interactive UI in the automated workflows. So when we look at goals-driven prioritization, like, what is that change. So that manual prioritization selections teams are driven by targets. And when we look at the prioritization techniques associated with those collections. Those targets are very individualistic and it doesn’t really drive to the ultimate goal, say, a CFO or an organization around DSO. So what we’re trying to reduce deductions, show hands. How many people have collection agents that are really good, really efficient. But at the same time may not be driving the impact to the DSO because the KPIs don’t ultimately lead to the top of the house in line there. I’m assuming that that’s the case right now. Well, when we look at that, goals driven prioritization. Well, we have the information. We have a behavior history where we can put that artificial intelligence to the goals driven prioritization, where our system is able to recommend appropriate actions that drive the biggest outcomes to do so. So the one team one dream. That’s essentially what we’re trying to ultimately get to is to align with what the executives are trying to achieve. So payment behavior based prioritization are essentially what we want to do is imply no customer is going to pay six days past due every month. I shouldn’t put any effort towards that. Because I know what’s going to happen where I should be focusing my time and energy is on that low predictability. The low predictability is what should be scaring you, because we don’t have any historic goals to be based on. So when we look at our AI model, we’ll look through the customer profile attributes, our history, past behaviors and past disputes. But then when that AI based behavior comes into play, we start focusing on the buckets of high predictability and low predictability. And what you ultimately get to are these insights of amounts and predictability associated to DSO impact. So. What we typically see today is. The top information your standard total customers number of collectors ratio total outstanding this top section. But imagine being a collector going into their workplace today where they have a bifurcation, a path to overview and current invoices collections of firefighting activity. That worklist is always continuously trying to get smaller and smaller and smaller, but it never does. But at the same time, there are capabilities to be able to go on the offensive that. Let me do some follow-ups with current my current invoices that are going to drive a large DSO impact. So when I look at a DSO impact with this screen and showing is that for past due overview. I have 910 customers would pass to but if I focus my effort on 138 customers, I can reduce my DSO by 5.6 days. Same can be said in the current overview. Within that collections, I can focus my time of 37 customers and reduce my DSO by 2.3 based off those payment histories. Now as we get there, being able to categorize what those subcategories are, low predictability, disputes, and then within each bucket you also have where should I focus my time and effort? Because again, the theme is DSO. So where are we today? How does the Thomas view of this? We turn things over to Liza. [00:05:00][295.1]

Liza Mohanty: [05:03]
So now that you have the list of customers that you need to work on, let’s see how that’s going. The journey of the collector is going to be transformed into autonomous world. So during the collective spend time on internal review, as they do review meetings, they do internal collection calls, they respond to inbound emails. So while we have very less self-control over internal things, we can bring in impact as the calls that are making their mom calls are coming to the system, the emails sending out and the incoming emails are coming to the system. And this is a day in the life of a collector. And of course, we have just simplified everything. And the call an email journey , much like the B2C revolution in the past decade with consumer applications like Netflix, uber . But we believe that autonomous is the way forward in order to cash flow. So our new model of collections is supplemented by several path breaking, autonomous communication and collaboration capabilities, which have been designed to have collective drive not only effective but also efficient interactions across mediums like voice, like emails and sms. And I just walk you through how that’s how we’re going to transform experience for the collection and use. So based on analysis of the date, she sees a weekly target assigned by the Monitor, which is now 1.24 million. She is 27% complete. So the way the system is predicting should be able to collect 900k through the week. I have calls to be made what remains to be seen and I have the data. So impact what you said Georgia spoke about. I have an overall portfolio summary that talks about what we do, current user overall information about the portfolio of customers that I have. Freeda has not only kept the calls and emails to the Senate, it has actually passed through all the incoming emails as well, the requests that have come in for invoice copies or account statements and actually created those actions in the worklist. So today, when I collect a second, you really have to gather all customer details manually from different sources. They call the customer manually. They, of course, that can refer to applications for extra information. They create actions manually. They will have to save notes, send out summary emails, how that’s going to transform in the autonomous world. And this one is we have a customer 360 degree view, which will give you a quick snapshot of the customers talking about what is total due amount , what’s a credit utilization, what are the payments have come in for the customer. So with all of that in mind, you make a call. So our integrated in-app calling, you still have easy access to all the customer information like invoice details. And if you want to access a new customer communication history that happened with the customer and freeda is there to transcribe the call, it summarizes the conversation, creates actions, emails, gets sent by freeda. So we basically feed the entire end to end calls, you have to use automated and even one. Today you send out emails, but when you send out an email, it’s like a black box for you. You really do not know what happens to be able to give the customers that actually engaging with it, reading through your emails, responding to your emails or not. And then if you do not respond, you manually follow up with them and then they might come back asking you send me the invoice copies or send me accounts statements. You collect is, you know, download the copies from the system, send it out. And I have not responded to any particular email that are repeated follow ups that so how this is transforming the autonomous world that on this one again is we have a small scheduling option, but you can small schedule an email at the right time depending upon the behavior of different customers and what time to read the emails that it will be delivered at the right time. So it would serve email a reading rate of what it is that you’re sending out. And whatever emails we are sending it will have interactive elements like they can create from is to be. They can create dispute within the email inbox itself without having to move out and access another application, the network. You know how that works and the actions would be created in real time collect to get to real time notification. We also are tracking emails and sending automated contextualized responses. The different examples that you a customer sees that he’s not available for the next two weeks. And there is some conversation that have happened when the follow up happens. We actually create a personalized response when you’re sending it saying an email next time and when your customers are actually sending requests for invoice copies, account statements , freeda actually passes through your emails, gives all the email responses ready, and invoices attached the collector has to just click on send and that’s it. So this is the dashboard I just spoke about and please feel free to drop in my autonomous demos video. This is all live. You can experience this that. So if I click on emails to be sent messages Im not too excited. This is my work list, but it’s already prioritized because based on the parameters that are set, if I for example, I have to send a finding payment reminder to this customer. Now, it has got two features that I spoke about, and this marks an option that I spoke about. You still want to talk standard. You have them, you can click on send that are interactive elements that create P2P , create disputes, schedule callback embedded within the email. Let’s see when next week on saying how it appears to your end customer. So your customer will have a view like this that an email has come from lara . open the invoices. And there are actions likely to be fewer disputes scheduled to call. But in the email inbox itself, when she plays some free to be the invoices list open up, they can do select. So you can select the invoices. Select the date that you want to make payment on and click on Save. When P2P is created at your end, customers then not get a notification, not autonomous, that a payment has come to be has come into lose the model from his or not take sides and do this. You must be sending out a lot of emails. Let’s say if you have a customer who has not been reading your emails for the last 30 days. So if you will actually identify such customers and bring it to your watch list, for example, when I get into grade one, it’s telling me in the suggested actions you call this customer because I’ve even sent in the last 30 days are not read. So when I go to my communication history, I will have track and status which events are rate, which are not paid. If there was a payment linked into that email, if that link was clicked, clicked or not. Based on our interactions with lot of customers, what we realize is 20 to 30% of time is spent on sending out responses to incoming emails. And some customers do have institutions work to respond to the emails. So a few would have actually passed on the emails and created actions. For example, if I click on send voice copies, I will go into the customer gene that be. I’ll be able to see what’s incoming email. What’s going on has requested an invoice copy from March 2022. The response is also drafted. Thank losses, copies digitized. You just going to click send your down. Similarly, if somebody has sent an email saying he will make a payment in another two weeks because my manager is not available due date. So that intent is also captured by Don Imus. You will have a quick snapshot view. Our system confidence score, for example, it says it’s 95%. If you are still looking at that, you can just approve it. Yet it an the pages get created and this is still if you still want to get into the region, you can click on the customer card. You will get to see what the email is about. It has said the payment will be processed by 30th April. So this action, this opportunity to do test capture the amount date who has committed the payment, just need to click on save and the fee is created. You click on reply, you send some others to summarization. Do we have customers payment coming received? But this amount to be paid by this due date. So yeah, so that’s about it. And this is on night in the product, like you mentioned. Please feel free to drop by the autonomous demo stations and I can walk you through this. You can actually be yourself. So I have to summarize. However, each line slowing the process of collections is that as a collector, I need to know three things. Then what is my target? Who should I contact? How should I contact to address that? We have a block list which you spoke about that considers two important factors that is which is of utmost importance to the CFO and customer’s payment behavior. And once I know which customers I have to work with to meet my goals, I have a I’m doing optimization of the course of events that will enable me to touch upon more customers and ultimately help me in my power collection metrics like CPI or Verso. So our number of calls, the whatever targets you have set for you.

Speaker 1: [14:16]
Thank you.