This is Doug Green and I'm the publisher of Telecom Reseller and I'm very pleased to have with us for the first time, Greg Whiteside, who's the co -founder and CEO of Human First. Greg, thank you for joining us today. Thanks, Doug. It's a pleasure to be on your show. Well, as I was just mentioning in our run up to recording this podcast, it's very rare for me to be able to do a podcast where the word human is anywhere in the title of even the podcast or the title of the company or anywhere. So that's a kind of refreshing new things. You're a relatively new company. You're in the AI field, but I think you're doing something very interesting and exciting in the AI world. So we're going to be jumping in on that and looking at that in just a second. But first, what is Human First? Human First is a data productivity suite for text and conversational data. So we help teams make sense of large amounts of unstructured data and turn it into very actionable insights that help drive their product and strategy, and also build AI projects faster by leveraging the data that they have. Now, you said something I haven't heard before in our discussion before we started our podcast, the search for use case zero. What's that all about? So I was talking about use case zero in the context of large language models and specifically how enterprise can leverage large models today in repeatable use cases. So since the since chat GPT came out, Human First, like all companies, looked at the impact that it would have on us and where things were moving. We had a lot of great insights already in terms of the conversational AI, and in general, text AI, and started really trying to understand what is the most repeatable high value use case that we believe any enterprise customer can apply large language models to today. We've heard from our customers and from a lot of other companies that this is a very high priority for them, their boards, you know, to show that they're embracing and that they're leveraging this technology. But a lot of the use cases that we saw were very experimental and hard to, you know, hard to do in a repeatable way. So we've been really focused on making sure that we understand what is the most repeatable high value use case. And this use case zero, we think, is really around using large language models today to make sense of large amounts of unstructured data in ways that weren't possible before with technologies like natural language understanding or, you know, even even more basic than that keyword search or semantic search. So you were telling me that your your major customers consist actually of enterprises, is that right? Yep, that's right. So we work with customers across all verticals from financial sector to medical sectors to telcos. And we work also with a lot of consulting management companies and agencies. So Greg, why did they turn to you? So our customers all share the same problem. They're interested in building and improving their customer experience with automations with AI. Some of them have deployed and productized some products, and they reach a certain point where they realize that the AI models are not the limiting factor anymore. It's the quality of the data that they're preparing and that they're using to train the AI that becomes a bottleneck. So our tool helps those teams really work efficiently on that unstructured data and make it very useful for training AI. Now with large language models, what we're understanding as well is that it's not just about training AI and building automation. If you can help organizations make better prioritization calls around what should be automated, what are the problems that can be solved with AI, and what is the best way to tackle those problems, you can see even bigger efficiency gains. So we're helping teams not only improve the quality and the speed with which they develop AI, but also make better decisions in terms of what should be automated by starting from the ground truth, which is in their voice of the customer and other conversational channels. So Greg, could you give me an example of maybe where you started working with an enterprise customer and sort of before and after kind of thing? Yeah, absolutely. So one of our customers is one of the largest last mile delivery companies, not only Canada, they're also working in the United States. So they're very strong partners with Amazon, do thousands of deliveries every day, and their contact center staffed by humans and with very little automation. So they came to us because they know that in certain periods of the year, there's a lot of influx of calls, and ultimately, they want to improve the customer experience and the automation levels. But they weren't sure, like a lot of call centers out there, exactly what the problems are, but mostly be able to prove out almost the business case for those and the ROI before even starting the project. So what we saw is they really wanted to have a data driven kind of approach to identifying what are the top opportunities for automation or product improvements, because it's not all about AI, it's also about identifying, you know, opportunities within the operations or product itself. So with Human First, we're able to ingest all of their contacts and all their call data, and very, very quickly build a very custom taxonomy of the call drivers, but also more deeply than that, the resolutions, actions taken by agents within the calls, you know, with the use of large language models, which allow us to do this analysis at a higher level than, you know, the raw unstructured data itself. And this allowed us to bring, you know, to surface some really, really key insights around some major blockers or friction points that affected, you know, over 30 % of the calls that they had, that with simple automation that we can show the functionality of, because we have all the flows and the edge cases in the conversations to show how you would solve it, you know, leads to a very big reduction in terms of time spent by the agents. So this is the type of project where, you know, with the right data -driven tools like Human First, you can start from the data and look for opportunities or problems to solve. And we did this really successfully with this last mile delivery company, and they're currently automating those flows that we brought to them and expect, you know, millions in ROI from that work. Now, was there an impact, let's go a little bit deeper on this, was there an impact on employee experience? Let's start there on EX. Yeah, so to be transparent, so the part of the project and what we bring is really this data -driven decision engine saying, here are the opportunities, here's the detailed, you know, analysis showing what are the different ways that you're going to need to be able to automate this particular, you know, within the contact center platform, which happens to be Amazon. So I'll be able to report on the, you know, end user experience from that particular project probably within a few weeks. Right. And that'll be the CX part, the customer experience part. Exactly. Yeah. You know, stepping back from that specific example, though, it sounds like this is the human part, that the human being that called in, the human being that's actually taking the call, the idea basically is everybody's having a better experience due to the automation. Is that the idea? Everyone is having a better experience if the friction points that can be, you know, within that conversation avoided, you know, lead to higher quality interactions afterwards. So to give an example, for this customer, I was talking about a really big part of the conversations was about figuring out the customer's ID and validating the user. And there's really, you know, much easier ways to do that than via human conversation. But it doesn't mean that the rest of the conversation can't be human to human. In certain cases, it's necessary. So, you know, Greg, how does this offer value to the enterprise at the end of the day? We're seeing enterprise really need to look at their data under a very custom lens. What we're solving in a sense is that a lot of the a lot of the products out there that help companies make sense of their data are very top down in black box, in a sense, they're really favoring speed and simplicity, kind of like a one click, put in your data one click, and will give you visualizations and dashboards. What we know is that that's not very actionable. And the reason is that those dashboards and insights are very hard to tailor automatically to your specific business and needs. So with a tool like ours, the real value prop is that we help build an extremely custom view into what's happening within the organization. And that in turn helps drive really data driven decisions and identify opportunities that you might not even have known you had. And we're really going from a lens where AI is a tool, but not every problem should be solved by AI. In certain cases, improving the product itself or parts of the operation will have a much bigger impact than automating the customer's requests later on. So what ultimately we want to help companies do is almost replace the customer support by fixing identifying and fixing problems upstream. And that's really done when you're able to have such a very custom understanding of what's going on and the data to back up the solutions to solve them. So that was very interesting that the ROI eventually is really maybe in the finding out of something you weren't even looking for. Exactly. And I think that's where our tool is very agnostic to the use case you apply it to. You can apply it to explore data, to improve AI training data, and to apply exploration to different types of data. And it's true that I think the biggest ROI you can bring to an enterprise is to help them tap into something like that data that they were never leveraging before. And that has a multiplicative effect, I would say, within the organization. And organizations that we're working with are building this data practice of centralizing their data and of disseminating kind of the value across different projects. And I think it's hard sometimes to measure the direct ROI of that. But clearly, we believe that it's this data that holds a lot of value for enterprise companies moving forward. Well, Greg, I really want to thank you for joining us and giving us a first look at Human First, and an interesting look at doing AI a little bit differently and approaching this challenge in a very different way. Where can we learn more about Human First? Yeah, I invite you to come to our website, www .humanfirst .ai, reach out to our team, there's a contact us button. And yeah, we'll be very, very happy to run you through the platform and to talk about your needs. Well, I hope to hear more good things from Human First in the future. I hope we do get to do this again and get an update on what you guys are doing. But for now, I want to thank you for joining us today. Thanks, Doug. It was a real pleasure. Thank you for having me.