Blythe Brumleve:
0:05
Welcome into another episode of everything is logistics, a podcast for the thinkers in freight. I'm your host, Blythe Brumleve. Maybe we were proudly presented by SPI logistics and we've got a great show for you today. If you've listened to any portion of this podcast over the last year, you know we're low key, obsessed with all things AI. Now we get to actually talk to a startup founder within the freight space that is focusing on AI. So happy to welcome in Mike Zayonc. He is the co-founder of Kodif, an AI powered customer support automation platform. So, mike, welcome to the show. Long time no see.
Mike Zayonc:
0:42
Yeah, thanks. Well, I think we're. I saw you a few weeks ago, but good to see you again.
Blythe Brumleve:
0:47
Yes, it was, and I don't know if you remember this, but at my very first or the very first freight wave show ever, I was demoing my former product. I don't have that product anymore, but I was demoing a product on the innovation alley and I remember meeting you this was like four years ago and that was and it's cool to see like the other side circle.
Mike Zayonc:
1:12
Yeah, now I'm the one demoing products and innovation alley.
Blythe Brumleve:
1:17
Yeah, that's right. So. So we were both at freight waves at three conference recently and you being you were in the VC world and then you left the VC world to go into a working to startup. So what does before we get into you know, sort of Kodiff and that background and what you guys are working on now I'm curious as to what your role just I guess, plainly, how does one become a VC?
Mike Zayonc:
1:44
Yeah, so becoming a VC there's many different ways. Like it's a very competitive. Like in finance, I think venture capital is one of the hardest areas to get into. It's a lot of people from top business schools all want to work in VC and, yeah, I was just given an opportunity kind of more on the entrepreneurial side. A lot of people also become VCs who are startup founders and that's kind of where I came in and the founder of plug and play hired me and we were actually really small when I started, so I was a pretty early employee in plug and play. Like we had 15 employees when I was first starting there in 2015. And today plug and play has about 800 plus employees. So we grew a lot. Vc is a big part of it and that's what I was mostly doing towards the end of kind of like my full time involvement with plug and play. But, yeah, it was. It was an interesting journey for myself, like kind of going through plug and play, and plug and play is the most active early stage startup investor in Silicon Valley, so it makes over 200 investment startups a year. And yeah, I started the supply chain program, as you know. So that was kind of the fund and the program I grew at plug and play.
Blythe Brumleve:
3:02
I'm curious. I said why supply chain? Did you have a background in supply chain? Or was this, you know, just an opportunity or an industry that that was ripe for the picking, for innovation?
Mike Zayonc:
3:12
Yeah, so it was definitely the second part of it. So I yeah, I was an early employee in plug and play. I actually started off in the media vertical. So they had me looking at like media companies and trying to work with like the huge media conglomerates in LA, like Disney and the sports teams and like the NFL, and I kind of realized like these are really like cool things people want to be associated with. And then, after like kind of a couple years in plug and play, I saw where a lot of the value was being had was in the B2B sector. So plug and play had all these massive corporations, fortune 500 companies, companies from all over the world coming to Silicon Valley to meet startup companies, and a lot of the activity I felt where we could have a meaningful impact, where the companies were actually paying the for innovation and taking risks, was more in the B2B sector. So plug and play at the time had like we had like four verticals we were in fintech, we were in healthcare, we were in media and then we were in, I think was we started like more of an industrial vertical and then we started launching a bunch of other kind of platforms and we're in retail as well. And now plug and play has about 15 verticals and I saw the growth, especially of insurance. So we launched an insurance program in 2016. So like kind of like a year or so after I joined plug and play and it just exploded. Like all the biggest insurance companies join plug and play and and like like just can name like the big ones, like nationwide state farm, like they started looking at innovation with plug and play. And then I got the idea I'm like we have plug and play hasn't done very much in logistics. And then I we saw like all these companies like Flexport and project 44 and raise pretty large VC funding rounds and so that all kind of came together. And then I kind of came up with the idea like, hey, why don't I focus on supply chain? And side the founder, he's very entrepreneurial. He basically said, if you can build a supply chain and logistics program, like go and do it. So I basically reached out to a bunch of supply chain companies and we ended up signing Maresk as the first partner and they were an incredible partner because they really saw the whole supply chain. They're very respected in the industry. And then, after we signed them, we ended up signing about 70, like of the largest supply chain corporations in the world. Like every motive transportation, we ended up signing a lot of the largest shippers, like Walmart is a is a huge, actually one of the largest partners of plug and play today. So it was cool. Yeah, looking looking at the the end to end supply chain all the way from the mine to last model deliveries at the consumer's doorstep, was kind of the vision of the program. And since I started the program and it grew in 2017, now there was about $6 billion of venture capital in the market. In 2022, when we raised the fund, there was about 24. So it grew like crazy and we saw like logistics all of a sudden became like one of the hottest markets, especially in the midst of COVID. So that was a really cool kind of kind of a wave to be a part of and it kind of led to us raising the plug and play supply chain fund, which is a $25.5 million fund. So, yeah, that's what I've been doing the last few years. And then I just got inspired and the reason I became a co founder at Kodiff is I just became inspired by all these companies that went through plug and play like we had like flexport founders, project 44 founders you can fork it's like just name the hot logistics startup. They were involved in our program in some way and kind of getting engagement with the large like supply chain companies like Ryder or JB hunt, etc. And it was. It was really inspiring to see them grow. And then we ended up actually investing in a number of successful ones, including RAPI, which became like the first unicorn from Columbia and a huge platform in Latin America. We became an early investor in Einride, which is fully electric autonomous truck, and also Shippo was another plug and play company. So it was pretty cool to see those those big success stories. And and then, yeah, like I saw a number of also people on the VC side like I think we talked about, like Jason from project 44, like he, he did the same thing and I'm like it would be good for myself to get experience like I'm dealing with all these really smart founders, to actually be in the midst of it, right, and then I thought I would get a lot more respect among the founders as an investor. And then it's just it's cool, it's it's it's always fun being and growing a startup.
Blythe Brumleve:
8:13
Yeah, I imagine that, being on your side of things, you already know what's important from a business perspective, and so you were able to take some of those skill sets and apply them to being a founder versus it. Maybe it's kind of the reverse for a lot of folks they start off as a founder and then become, you know, an investor, whereas you you got to see it from both sides and now you're in you know the startup founder role. What were some of maybe those, I guess, key skills that you learned on the VC side that really help you on the founder side of things?
Mike Zayonc:
8:45
Yeah. So I think what I really understand well is like how to grow a VC funded business, because that's what plug and play does every day like and I think there's like it's kind of like a recipe in Silicon Valley like you hit these, these metrics of growth, like the VCs want you to grow fast, like they all want you to have like a really good team. You have to have those ingredients together. You have to be going after a big market. So from the VC side, I kind of know exactly what they want and like what we need to do. So that side now is is easy to understand. And then the hard part is actually doing it right. Because it's, it's easy to everyone can always like like poke it at an entrepreneur and say, oh, you're, you're, your, your market's not good or you're, you need to improve this or that, and the hardest part is really doing it. And I think building a successful startup is is about execution, and I think that's the hardest part. So, and yeah, that's what I'm excited to do today is just kind of like work with the code of team execute, build a great business. And it definitely has been very helpful because I I'm very aware of like the, the financing metrics we need to hit for each funding round, like how to position it, how to work with these corporates, because at plug and play, we used to like get feedback from our. Plug and play still does get feedback from all these corporates what do they actually think about the startup? So the startup would like pitch them and then plug and play. We'd work consultatively with those corporates and then if there's like not a match, for example, or they want something different, we could recommend another startup company. So it's been incredibly helpful to see this like full circle, and I think there's a lot of people that have gone from VC to startup and vice versa, and I think they really like help each other as well. So it's it's it's been extremely helpful for me and I'm extremely grateful to plug and play, because I I still am kind of like sourcing companies and recommending startups and, yeah, like I, I need a meeting like another freight wave startup learn to win. So they're like a Silicon Valley training platform and I was just chatting like in the Silicon, we're both logistics like startups in Silicon Valley. And then I referred him to plug and play and plug and play just invested. So nice. Yeah. So it's cool to like, yeah, kind of work full circle and also recommend startups as well. So plug and play has been very good to me and that's been a huge advantage of myself to kind of leverage that network I built from plug and play and kind of those relationships, because I think in logistics as well, like a lot of it's about relationships like and that part takes a lot of time, that's that's. It's not like hard, it's kind of like it's fun building those relationships, but it's if you're a new startup, that's something you have to like also consider as well.
Blythe Brumleve:
11:45
And so what was? I guess the, the, the catalyst to say, ok, kodaf, is, is the right startup that I'm going to jump into and that I'm going to invest my time and energy into.
Mike Zayonc:
11:56
Yeah, so I invest, or my team invested, in over 70 companies in the supply chain program and and, yeah, some, a lot of them have been extremely successful and it's a it's a very long term thing, right, like you, as a VC, you don't get an exit until there's an IPO or like an acquisition, which is normally like seven to 10 plus years. So, yeah, so I looked at various companies and wanted to evaluate which one I can maybe join or help more, and I got a lot of startups that would come to me and be like introduce me to this company, like I need an intro, like please, and I would constantly get these, these kind of request and and I helped a lot of startups and I would just be like, ok, like I'll make an intro for you, because plug and play that's that's part of our business model, is helping entrepreneurs and so that. So that was fun and like I mean we had a lot of meaningful introductions that that led to like large partnerships with a lot of the startups you know, and logistics like to like companies like Walmart signed with them, or Yamato or these huge logistics companies. And then Kodiff kind of asked me to do the same thing. I was connected to the founding team because there's like an old guy who worked at, who worked with me at, plug and play. They went to Stanford and one of my co founders was there and and there's, there's that relationship. So we all kind of like knew each other. And then I saw these guys and I'm like these guys are very smart, they have a great product, they have a great technical team. The thing they were, I think, lacking was like an access to the, the industry Like they. They were not like deep in the logistics industry. They were not deep with like these massive, like enterprise customers that I worked with in plug and play. And so I started making like some intros to them and I thought like, yeah, like as, as, as you make an intro, it's it's. You can't always just like make a referral. Sometimes it works, but it takes a lot of time also from the team to kind of build those relationships. So so, so, yeah, I just saw it as an opportunity where I could leverage some of those relationships I had with huge companies with within plug and play and then engage deeper. So that's, that's actually what I've been doing, so just started making intros. Now I'm full time kind of promoting Codiff, growing Codiff and and working with these, these huge corporations, to kind of like help them leverage generative AI automation in our platform. And and I think the other thing you have to factor in, like when I joined Codiff, is you have to get along with everyone, like there has to be a cultural fit. So I thought that culturally, like I just love the team, I like kind of their mindset and and yeah, like I got opportunities like over the years from many startups and well known startups and logistics to work with them and yeah, sometimes you think about it and some of them have became very successful. And yeah, I think, with, with with Codiff, like everything, everything there's a lot of boxes that needs to be checked before you make that jump. And they allowed me to come on board as a co founder, which is really exciting and so over overall definitely was not was not easy and yeah, I talked to some other startups as well and a few of them. There's always like one thing, maybe that didn't kind of work out and I mean I wasn't like in an urgent situation. I had a really great position in plug and play, so I could just keep kind of investing and the plug and play fun is still doing really well. So so, yeah, that's kind of how I ended up there.
Blythe Brumleve:
15:53
And I guess what was your background or experience with AI as before you joined Kodiff? Or has this been like a learning experience throughout that process? Because it really has just sort of I guess, quote unquote gone mainstream even though I don't feel like it's very mainstream yet is just the general awareness around AI and how it can be used in day to day. I'm familiar. What was your familiarity with AI? Were you using it? Do you kind of see it as sort of like the future? And that was a big catalyst for joining Kodiff.
Mike Zayonc:
16:28
Yeah as well. That's another reason, like the market was incredible, right? So everyone's now talking about generative AI, everyone's talking about automation, and I saw the thing that also really stood up to me with Kodiff was our CTO wrote the white paper at Uber to automate customer support. So that was the largest issuance of machine learning for any customer support application at the time. So this was like 2019 as Uber was going public, and I saw that and I'm like that's pretty incredible that he kind of had that foresight to use such a high and it was interesting that Uber like people love and hated it at the time that you literally couldn't contact customer support for a while in Uber and what they ended up doing they kept improving the product and now it's like providing full on refunds. There's a lot of like AI in the background that fully automates those requests. So right now, they're deflecting 70% to 80% of customer requests globally for Uber and Uber Eats, which is not an easy thing to do. So I like that. He was early in the market and had the capability to build a technology platform like that. And, yeah, everyone now is talking about generative AI. It's definitely by far the biggest trend I've seen in the startup market Like I've been through many different cycles, like IoT was a big trend in getting data, and then it was like crypto and blockchain and now, I think, meta person. Yeah, there's always these cycles right Of what's the hot area, but I think AI is way bigger than anything. That has become a big hype cycle and I think it's here to stay and I think that everyone is saying that this is real, this is transforming the market and you see the real business use cases in logistics and in every industry. So, yeah, that was another driver. I wanted to be kind of even deeper inside of the AI world and plug and play. I mean, we were investing in countless AI startups as well, so I got to see from the other side like how they were growing. And I mean, when I started the supply chain program, automation, I think, was always one of the biggest areas. Like we're always looking at automation tools for warehousing or how you can make different processes more efficient to optimize your supply chain. So that's something I was already looking at a lot inside of or inside of plug and play, and I thought it would be good to apply a lot of that kind of background to as well.
Blythe Brumleve:
19:17
Are you in freight sales with a book of business looking for a new home? Or perhaps you're a freight agent in need of a better partnership? These are the kinds of conversations we're exploring in our podcast interview series called the Freight Agent Trenches, sponsored by SPI Logistics. Now I can tell you all day that SPI is one of the most successful logistics firms in North America, who helps their agents with back office operations, such as admin, finance, it and sales. But I would much rather you hear it directly from SPI's freight agents themselves. And what better way to do that than by listening to the experienced freight agents tell their stories behind the how and the why they joined SPI? Hit the freight agent link in our show notes to listen to these conversations or, if you're ready to make the jump, visit spi3plcom. Well, let's get into some of those use cases for some of the products that Codiff has, specifically when it comes to logistics. Because customer support you just said it about Uber it's such a big, I guess time suck for in a budget hoarder, I think, for a lot of companies, especially in logistics, where I've always been told, for like 10 years now, if you can reduce the amount of phone calls that a brokerage gets or that a carrier gets, then you're increasing the efficiency and productivity of the workers for that company, allowing them the time to create better relationships, which is so important, and you hinted to that earlier. What are some of the products that you guys are creating at Codiff that addresses some of those sort of customer support issues through AI?
Mike Zayonc:
20:53
Yeah, definitely so. A lot of companies they'll outsource a lot of those repetitive calls and they'll find like what's the most cost effective way to kind of manage those requests. But in the future all those high volume transactional requests, they're going to be automated. And what's been interesting at Codiff is when we're working with some of these customers, like Good Eggs, for example, when they're like a grocery delivery company and their customer service agents are actually now happier as well. They don't even like answering those repetitive requests. So with Codiff we basically can take all these high volume repetitive requests that, like in logistics, the biggest one obviously is like where is my freight, where is my order? So that's what we were doing with these e-commerce companies early on with Codiff, where we would just immediately they would put in a chat bot when is my order? They would authenticate them and then they'd get their requests and normally that would take them a day or two or maybe the customer service agent would take like 15 minutes and eventually reply. So now that's automated in real time. So that's like one use case. But I think there's a lot of those high volume repetitive requests. Or you want to access a bill of lading, you need insurance proof, or there's a lot of other use cases. If it's like high volume and transactional, then those are the type of use cases that we're excited about.
Blythe Brumleve:
22:39
And so what I guess for the product's sake? Does it live on top of a customer's website? Is there a dashboard within Codiff? What does I guess sort of the how is the customer using your product?
Mike Zayonc:
22:55
Yeah, so we just sit on top of whatever infrastructure they have. So it's in the beginning. We started textual. Now we're doing every, we're moving to every mode of communication so phone calls, et cetera but for example, on the website, on the app, we can set up a self-service chat automation platform. So let's say you're a logistics company or you're an e-commerce company, we can just put Codiff on your website and then you can just type in chatGBT, for example, like any requests, and then you can automate that and have a smart AI that will resolve those issues. So that's been very effective and there's also now a lot of interest going to market with a lot of trucking and LTL and brokerages that there's an opportunity to automate those high volume requests as well. So, yeah, we basically can read any textual data is very easy, and the reason we started off with textual data and ingesting that is because it's more clear. When someone's calling in for example, like a lot of trucking people will call in to a representative if they're not available or they have to wait on hold, they may want to get an immediate response, especially if it's something like a lot of brokerages and trucking companies would constantly get just like where is my order request. So if you're waiting on hold for like 5, 10 minutes, you might just want to automate that process and just do two clicks with Codiff and it's done, your order is tracked, and so, yeah, we basically for self-service. We can sit on top of whatever platform you're using, whether it's an app, website and then we also what's interesting with Codiff if you don't want to have that full on self-service or you want to minimize the risk of using AI directly with your customers because it's a big jump as well to try to automate 20% to 50% of your request in the beginning with AI Then we also have agent assistant tools. So we have a Chrome plugin. It sits on top of any platform, so we can sit on top of your TMS, it can sit on top of your ERP, your ticketing system, your Salesforce account, and so it's essentially like a smart tool so you can type in any request as a customer service agent or some brokers are now using this as well with us, so you can type in those questions as like, let's say, you're a new customer support agent or you're a new broker and you don't really understand where to find this or how to find this. You have to go to your manager and then ask for these repetitive questions and then the manager gets frustrated because they're like, oh, I taught you this on day one and everyone kind of gets frustrated and so we can just automate that with like a chat GPT, like experience. And then the other issue is like a lot of customer support agents, when they're going outside of their organization and like giving these questions to like chat GPT, like they might, like they don't know how to respond, so they'll kind of give sensitive company information just the chat GPT which could. You don't know what's gonna happen with that. So this is like a very secure and safe process where it's streamlined, it's all based on your company information and we track and we provide analytics also in terms of, like what's actually happening in your organization. And so what's cool is like we can also do like a thumbs up, thumbs down each response and we track those responses. So our AI always improves. So if, for example, you wanna just like test out Codiff and you're like, okay, ai might be too much of a risk in the beginning, then you can just test out, like our agent kind of co-pilot product. Use that and your agents are gonna be 40% more productive in general, cause they're getting all these fast responses. And then you can use all the prompts, like for tracking. You can use like chat, gpt, like inquiries, like make the response friendlier, et cetera. Make it shorter, so that's really useful. And then you can even inquiry like things internally, like if you have a question on how to find something, so that's useful. And then once you're comfortable to use more self-service, then you can use that as well. And some companies wanna use both at the same time. So Codiff is like a holistic solution, so we basically can provide access to both, whatever the company is comfortable with.
Blythe Brumleve:
28:05
So it's almost like, if I'm understanding correctly, it's almost like cloning your management team, cloning your training docs so that your internal workforce is much more productive. But then also you have another product where it could be more customer facing, where the customer can get their answers much more quickly based on the database that the company is working with. Is that, I guess, a good assessment?
Mike Zayonc:
28:31
Yeah, that's great. Yeah, you can be a salesperson for Codiff.
Blythe Brumleve:
28:35
Well, I do wonder. So what is, what are some of the? So you mentioned TMS. You mentioned, I imagine, some of the internal processes, some of the internal work documents that these companies have, what does, sort of, I guess, the onboarding process look like? What kind of data are you collecting that helps for these companies to take that step towards AI?
Mike Zayonc:
29:00
Yeah, so any textual data we can read. So, yeah, like tagging into your yeah, like TMS, your knowledge base, wherever you're storing that data and it's definitely important to garbage in, garbage out so you need to keep track of your data in a proper way and so we'll basically bring together all those different systems that are important for you to leverage and it's an opportunity to to kind of like streamline that process. And yeah, I mean it will take us like normally we move very quickly, like we're on the startup side, we're very agile, so we can, we could literally we built chatbots for people in like five minutes. Like I did that at a few conferences and like we just tagged into their knowledge base and like built, sent them a generative AI chatbot. But obviously, like we want to go deeper and we want to like provide, like fine tune the chatbots, the resolutions, the whole platform. So normally it takes like to implement it and get it started, like to do a pilot. It will take like a couple of weeks to like get all the specific data and work with the teams within the organization. But, yeah, we literally, yeah, I can do it in five minutes If, if a company just wants to like test out or play with our chatbot and we built like this like playground, so companies, also online could just go and like build a quick chatbot and like test out the generative AI capabilities.
Blythe Brumleve:
30:37
What about and you mentioned something with you know garbage in, garbage out. As far as your data is structured, Are there any like common threads or common themes that messes up a company's data, that kind of slows them down or maybe even prohibits them from using an AI tool?
Mike Zayonc:
30:54
Yeah, I would say like generally, like I'd say the biggest hurdle that we have these days working with companies is like it's new, right, and everyone's getting comfortable with it. But I think those things are all important. Everyone knows they need to get clear data, they need to make it in a way that can be read by AI and that's, that's the future. So I think now it's like we're in kind of like this strategic point where we had a lot of adoption with kind of mid-market customers, so a lot of those customers were willing to kind of take the risk. But then when you're talking to like these huge companies, then there's there's going to be like all these engineers. Everyone internally is like oh, I can build this, like for for us, right? So there's always that build versus buy question right now. So I think that's going to be the biggest. That's actually I think the biggest slowdown for us is people are not questioning like do they need this or is this going to be the future? I think everyone recognizes it. It's just like how they're going to do it. And especially these huge companies, they're all moving pretty slowly like more at the enterprise level, like the just think about like the JB, hans, or these huge, like huge organizations. But I think in reality, like I think it's been proven like the reason I'm excited to do this on the startup side is like I think startups will provide great solutions here and, like, whether it's Codiff or other companies there, there's this application layer, where generative AI is now coming in and automating a lot of those like specific kind of workflows or areas with the customer. So it's like who's going to kind of get in working with these enterprises and build, to deliver the value and do it well? Because I think these will be the successful companies when you build a reputation like you've done this successfully over and over again. And that's why I also chose Codiff, because I knew like our engineers are from Uber and Amazon and they are very strong at kind of building these capabilities and they also had entrepreneurial backgrounds. But it's definitely not like super easy. Like you can't always just like tag into like chat gbt, for example, and build a simple like enterprise solution. So there's a lot more complexity and so, yeah, I think it's important to kind of partner with a company like us who has experience, does this every single day and and is very focused on AI as well. So there's also a lot of like like other companies, like ticketing systems, trying to build AI tools or tech companies. But I think, like, if you're very focused on AI as a priority, like working directly with like a company that does AI, this is our, our model and this is what we believe in, and I think it's a lot of those companies are also going to have innovators dilemmas as well. Right, like they, the more AI they have, like, if you're a ticketing system, they're going to have to cannibalize some of their business model because they're charging based on seats. So, like, if you're like Salesforce or Zendesk, then you are going to charge based on how many users are on your platform, whereas codiff, like, we're automating the a lot of those usage. Right, so it's the. I think the business models are going to change, so they're going to be more based around like resolutions and kind of outcomes overall, rather than just like traditional, like SAS models or just like how many seats your people are using the platform.
Blythe Brumleve:
35:01
So if I'm a company like you know, an SMB maybe not enterprise level yet, because I feel like enterprise is just that I mean even medium sized businesses and freight have so much data, if I'm thinking about, you know, my 2024 plans, I want to start the groundwork of you know, I guess gathering all of my data sources. How should I start preparing my business now in order to be able to use a service like codiff?
Mike Zayonc:
35:30
Yeah, so I would. I would definitely try to partner with companies that are able to gather data in a clear way that is able to be read like in a textual format and, like a lot of companies and logistics are pretty old school, like using, like mainframes and like the, so enabling also like API. So if you can enable like an API for like codiff or another AI company to tag into that system, that would be kind of like the requirement for us to read like all that textual data.
Blythe Brumleve:
36:09
So it wouldn't be something like. You know, I swear by Otter that the AI note taking app, so it wouldn't be something like that, or would that be a component to your data set?
Mike Zayonc:
36:21
Yeah, otter is awesome. They're very useful. I've used Otter, everyone loves it. Like you can record like all these conversations and go back, and I think that that's also a major trend like recording all that textual data, because because, yeah, if you recorded an Otter, you clarify in like a some sort of structured way all that, all that data. Then the AI can just read through all those conversations so it will be able to make like quick decisions in the future based on what's actually been said. So I think that's a that's tools like that are going to be very, very important as well.
Blythe Brumleve:
37:02
So a business owners probably need to start, you know, recording all of your calls, maybe recording some of your sales meetings, client meetings, but then also, for that next level, creating having your, your dev team, create an API and connect it with So, then that way you can start. I would imagine scraping some of that or just plugging into a lot of that information so that you can start bit laying that groundwork of being able to automate some of your customer support and things like that.
Mike Zayonc:
37:31
Yeah, exactly, I would recommend that, like, we focus mostly on the mid market in the beginning, but I think all layers are going to be able to use this in the future. So, if you're like an SM, smb, they're like we had a bunch of like kind of like mom and pop or like smaller businesses, like small e-commerce companies, that were looking to leverage us, but I think at this point, the ROI isn't always there as well, right? So it's like because we're a startup, so we need to also grow and be smart with our time. Like, if we found like a flower shop with like one specific area and just spent all our time doing that, then that could cannibalize a lot of our time on the business, right. So I think as the models get more developed, then a lot of these companies are also just going to be able to use it because, like, if you're like a small business owner and you just want to like automate, like all these repetitive, annoying questions, like, everyone sees value in that, right. So I think that's going to happen as well. Like it hasn't been our focus in the beginning, but ideally, like we'd like to start in the mid market and then move to enterprise and then enables a product like that self serve, like we can already enable customers to build very quick chatbots as well, but obviously like it's hard as a startup to spend too much time on that. But I would agree that any textual data is going to be the future. If you can get that recorded somewhere, then AI will be able to read it eventually.
Blythe Brumleve:
39:14
They start laying the groundwork now. But then what does that so say you have a good fit for a business to start purchasing Codiff or start engaging in that conversation. What does the onboarding process look like versus just the ongoing management? Because I imagine that there's going to be someone within the company that's going to be tasked with like the management of this, the reinforcement, learning and things like that. What does that process look like?
Mike Zayonc:
39:43
Yeah, so that's kind of. I think every company is a little bit different, so we should understand your workflows more. I think one of the beauties of Kodiff as well is that and what I loved about what our CTO built for Uber and the platform he scaled globally for Uber that's still running for their customer support. It was a low-code platform, so you had minimal kind of technical requirements. Like you could get kind of like a data scientist or one person who could just kind of like manage it and they didn't have to be like a full-on engineer. So we actually can enable these capabilities with zero engineers and so you can have someone that's like slightly technical, that can kind of just oversee the platform. I do think it's valuable to have engineers as well internally that we can work with, because there needs to be some people that oversee it and they'll probably understand it at a deeper level. So I think that's been a sweet spot for us is to find those companies that they're not like Google or like Microsoft or like they're like a tech company full of engineers, but they might have a few people that are pretty technical that can kind of leverage this and then they could just build from the kind of the low-code platform that we have.
Blythe Brumleve:
41:16
So we've talked a lot about the AI technology and automation technology that's here today. Where do you think some of these applications are going? What does it sort of it's tough to tell, I know with the future of AI, but what do you see as far as use cases in the near future, or maybe later, in five years or so, which I think is probably impossible to tell?
Mike Zayonc:
41:40
Yeah, well, I mean, when we're at FreightWaves, like Brad Jacobs was saying, like logistics is going to get automated, and that was cool to hear. I think like a lot of like I'm for sure these like highly repetitive questions that, yeah, like customer support agents are just, I mean, a lot of them are paid like a minimal salary. It's not a job that a lot of them want to do forever, like they might want to grow in their careers. Those are going to get automated and I think like there's going to be like a higher touch. So if you want to like speak to someone very consultatively, then you'll have that ability like more of like a premium level, and then the rest, I think, is going to get automated and I think a lot of those processes are going to get smarter and smarter. Like there's going to be a lot of stuff inside of different brokerages and logistics that are going to get smarter and smarter, like, like they always say, like okay, well, the broker knows how to find these specific orders and shipments, but in reality, like now, this AI like with generative AI, is becoming much smarter, right, so the concern will be around like the hallucinations as well, like sometimes the AI will make a mistake or you know that could be very dangerous. So I think doing it in a very safe way that will also not offend like your customers or kind of ruin anything with the business is important. So that's where you have to be really careful with like generative AI because, like, if you ask like even as like a Data-day consumer, if you ask chat, gbt Certain questions, like it's gonna make mistakes, and like when you're dealing with like business or higher value Items, then you can't you can't afford that risk, right. So I think for us, like, if we can Use codiff and like generative AI to learn and figure out what those responses are that are like, then it's like we eventually will be like much more accurate than than a human. So I think I think the future is going to be way more automated and, yeah, definitely gonna gonna change Significantly like a lot of those repetitive kind of workflows.
Blythe Brumleve:
44:10
Yeah, I agree, I think some of the the, I guess, lower value work. That's time-consuming, it's just gonna be automated away but it's probably gonna take a little bit of a process of learning, reinforcement, human learning, you know, against some of these hallucinations to get it there. I I use it pretty much every day between Claude and chat GPT, using both of those tools daily. Just by using you know transcripts from conversations like this, but it will still hallucinate from time to time, but it gets me to about 80 percent there and that Increased in productivity. For, you know, a small team like me is is a game changer. So I think that you know it's just the the start of when I think a lot of logistics companies can can take full advantage of what some of these tools are being offered or what some of these solutions are being offered. So, mike, where can, where can folks you know follow more of your work? You know, sign up for maybe a demo of Codiff? You know all that good stuff.
Mike Zayonc:
45:09
Yeah, definitely so. Yeah, I'm easy to reach. You can reach out to me on LinkedIn or you can just email me my emails. Mike at codiffai. Our website is codiffai, and yeah, I would love to chat more with anyone that Would love to learn more about our our platform.
Blythe Brumleve:
45:28
Yes, heck. Yeah, that was awesome. I Hope you enjoyed this episode of everything is logistics, a podcast for the thinkers in freight, telling the stories behind how your favorite stuff and people get from point A to B. Subscribe to the show, sign up for our newsletter and follow our socials over at everything is logistics calm. And in addition to the podcast, I also wanted to let y'all know about another company I operate and that's digital dispatch, where we help you build a better website. Now, a lot of the times, we hand this task of building a new website or refreshing a current one off to a co-worker's child, a Neighbor down the street or stranger around the world, where you probably spend more time explaining the freight industry than it takes to actually build the Dang website. Well, that doesn't happen at digital dispatch. We've been building online since 2009, but we're also early adopters of AI Automation and other website tactics that help your company to be a central place to pull in all of your social media posts, recruit new employees and give potential customers a glimpse into how you operate your business. Our new website builds start as low as $1500, along with ongoing website management, maintenance and updates starting at $90 a month, plus some bonus freight marketing and sales content Similar to what you hear on the podcast. You can watch a quick explainer video over on digital dispatch I oh just check out the pricing page once you arrive and you can see how we can build your digital ecosystem on a strong foundation. Until then, I hope you enjoyed this episode. I'll see you all real soon and go Jags.
