Building an API for Global News

[00:00:00] Mike Bifulco: Hello and welcome back to APIs you won't hate. I, of course, am Mike Biko, one of the founders of APIs you Won't Hate, and your podcast host. I am sitting down today for a conversation with a new friend of mine to talk through what is a really interesting API product with kind of a fascinating and unique story from inception to product two, where they're today.

[00:00:20] And I'm really excited to sit down and talk to Artem Bora from News Catcher. Artem is a, like me, a yc founder and unlike me, has a really, really interesting story of the inception of his company and how he's really built something interesting through many, many interesting little bumps along the way.

[00:00:39] And I'm really excited to get to chat with you. Artem. Thank you so much for joining. I, I appreciate it. How are you doing today?

[00:00:44] Artem Bugara (NewsCatcher, CEO): Hey, Mike. Thanks for having me. Yeah, I'm doing, I'm doing great. Happy to, happy to share something I've learned so far as a first time founder. Yeah.

[00:00:54] Mike Bifulco: So let's, let's start I think with, with Tell the World what let's tell the world what News Catcher is. So [00:01:00] what's your elevator pitch for news catcher?

[00:01:02] Artem Bugara (NewsCatcher, CEO): It is finding my elevator pitch for new sketcher changes almost every week, but I think, I think the latest one, I think the latest one based on like all the latest interactions and changes in the in the world is basically new sketcher. Think our assumption is, I'm gonna start with a problem, right? So my assumption is that companies don't really leverage information that can be found in news.

[00:01:28] Surely if you ask and, and, and the most important thing is like, I think they do, but they don't. But they actually don't. If you ask like. The company, they're gonna obviously say like, yeah, CEO management reads like Bloomberg or FT, or something else. And obviously they like really looking into how like Trump Biden debate is going.

[00:01:49] Like who's gonna win? How the interest rate gonna change, what the corporate tax, of course that's, and, and that's like important for big, it's important like as, as a big microeconomic shift, like as, as a big thing [00:02:00] going on, or like the war and conflict. But in reality, like the, that's not what usually makes company any like.

[00:02:07] Money usually in a day-to-Day business. There is like much more like direct and applicable things that like are important and a lot of those actually can be automated or like semi-automated. And so our idea is actually for all the, at least the largest companies, like each department working for like each industry in each region.

[00:02:32] Usually can benefit a lot by either losing less money or gaining more money by like being able, able to leverage stuff that can be publicly and openly available, not just like a big kind of things. So that's where we that, and that's where we, that's where we are today. But in more simple, in more simple terms, I would say that new sketcher basically.

[00:02:54] Provides a news, API, we aggregate publicly available news articles from over the, over, all over the [00:03:00] globe. And not just like big media, but local news industry specific news. And then what we do is we basically take this, we have over 1 billion or one and a half billion articles indexed in our database.

[00:03:12] All like index and, and, and easily and being easy to search. And what we do is we kind of try to meet an enterprise customer. Deeply understand what they need and then say like, look, we have the data that you kind of, that has insights and that has stuff that is important for you, but more of that we are gonna help you really like integrate it.

[00:03:36] We are gonna do like a bespoke integration. So a little bit like paler like approach. And I think that's our, that's our, that's kind of our thing. I think one. Over the last four years doing new Sketch, I think the most important insight we realized is that the biggest problem is not like, you know, getting data and like getting articles through the same shape and form from like over 70,000 [00:04:00] new sources.

[00:04:01] It's rather to be able to actually integrate it into like workflow into day-to-day of, business of some of, of some specific department in a specific, in a specific company. So yeah, that's, that's kind of where we are where we are right now.

[00:04:18] Mike Bifulco: When you and I first chatted, I was really interested diving in on your website, looking at some of the information on there. The, the heading of your website is something like news intelligence at a global scale or something like that. And from the sounds of it, so you've got a database with hundreds of thousands, millions, billions articles 70,000 sources and all focused on aggregating and providing intelligence around news.

[00:04:39] Artem Bugara (NewsCatcher, CEO): Yeah.

[00:04:39] Mike Bifulco: One of the first questions that my mind ran into is like, how does that, how do we get there? How does that start? That is a, a massive undertaking of a project, and I'm really curious about like the story that got you there. So can you tell me a little bit about maybe your career leading up to News Catcher and the reason that, that this came into existence?

[00:04:57] Artem Bugara (NewsCatcher, CEO): Sure. So basically News sketcher [00:05:00] has two founders me and Maxim. Maxim is CTO, I'm the CEO, but originally we are both non CAS degree data geeks. Kind

[00:05:10] Mike Bifulco: I like it.

[00:05:12] Artem Bugara (NewsCatcher, CEO): yeah. So, and, and probably like, I'm gonna start with like, so I know Maxim already for like 15 ish years, since like schooling back in Ukraine in the next.

[00:05:21] Then we actually studied together in one university in Ukraine. Then we got accepted to the same university in Strasburg. And like, so yeah, we got a lot of actually years been studying together. And then just for the few years, I think we've been like a little bit apart from like doing different master's degrees and having different our first jobs basically, but originally.

[00:05:42] So we got into, so like I said, we had like no CS degree engineers, so we like, we had to do all of it ourselves. And it actually started one day in Strasburg, France, where Maxim came to me and he said, I. I just like learn about this, [00:06:00] like Titanic dataset, Titanic, like machine learning competition on kale.

[00:06:05] Kale is a website for machine learning people to like compete with each other. And it was like, I think 10 maybe, I think, yeah, I think it was 10, kind of 10 or nine or eight years ago. And Maxim and I both kind of, we studied eh, we studied economics and I think economics was the last part we actually enjoyed.

[00:06:25] We enjoyed the most like mathematics and statistics and like game theory and probability and all this kind of stuff. And so like these Kaggle and like data science sounded like a nice trajectory for us to, to keep working on. And so we, that's how we got into like programming by. Learning RI think we, I think we actually,

[00:06:47] Mike Bifulco: Wow.

[00:06:48] Artem Bugara (NewsCatcher, CEO): yeah, I think we actually like studied on data camp.

[00:06:51] I think there was the website and I actually remember Maxim one day was like waking, I think he, he, at some point he took all the courses, all the courses and, [00:07:00] and, and I dunno how much actually he learned from them, but he was like really dedicated to like. Studied and doing something with that. So yeah, we, we've been basically slowly just getting into this.

[00:07:10] And for our master's degree, I went to Toulu School of Economics which is pretty solid economic school where the, I think the dean or the head of it has like a Nobel Prize in economics and Jeanti Hall, and I started the econometrics. And that's actually like, you know, like a, like a, like a more like applicable applied.

[00:07:32] Economic economics economic, like economic statistics mix. And Maxim actually started like something actually that was called, I guess, data science in Paris. But yeah we've been there and one thing that happened, I went to, after the first year at. Doing my masters, I found an internship in a French company.

[00:07:52] So basically the CEO is a TSE to School of Economics alum in Dubai, a vendor. And, and basically I was like an, [00:08:00] an intern there and then I said like, look, I don't want to really go back to, to university. I know how many more years of education I need to actually to actually work. I, I always, I always thought about it to be fairly honest with you.

[00:08:11] And so I started with my very first job where I was like, just like, I think I had a lot of freedom and I could like, try to improve a lot. And I think that's where like I understood that like, I really like coding and one day I, I had a, I had to do some web scraping job and that's how I said like, okay, it's finally time to like open Python.

[00:08:30] And I think this, in one day I started doing something in Python and then I came back to r. And yeah, and since then I, I was like really passionate about Python and I was trying to do a lot of like side projects as well as Maxim and we chatted about it. We really, we really liked it and I. I guess at some point I realized one thing, I don't know anything about Python.

[00:08:59] [00:09:00] I know like Python, but I don't even know what server is, how internet works. And so I had some side project where I decided to go from zero to one, like following like Y Cator startup school thing. And so basically I, I, I've launched like a, like I was like scraping some news actually, like comparing some coverage for, for the, some something in Ukraine.

[00:09:22] I think the, the, the presidential election. But basically it was like pretty cool 'cause I, I had to learn everything and I think I still remember this evening where I realized that, well, database is actually like, you can. Get it running with one Linux call. Like, it's like Postgres apparently doesn't have to be like, it's not a big building with like data.

[00:09:44] It, it is pretty much, it's pretty much easier than that. And I think that that was like, like really I was working a lot and like I was doing my side project a lot and for like two months I, I kind of learned a lot of stuff about like how things work, how servers work, what [00:10:00] AWS is like, how, how everything works.

[00:10:03] And I really kind of enjoyed it. And since then I felt like. It is pretty cool. I should probably do something I should probably do something something of, of my own with that. And I guess by the end of 2019 I was just looking for another like side project technically. Interesting. I just found this like other new EPI and I and I thought like, look, it sounds like a good idea.

[00:10:24] Even though I don't really have any problem with it. I don't know anyone who needs it, but I just thought like. It's a really cool technical challenge. And so that was like a side new sketch. It was like a side project in the, in the beginning. And so I guess if you, if if there's anyone listening to me like it's having a cool technical challenge is a, is a great reason for the side project and it's probably like the worst reason for like starting actually a company.

[00:10:52] If you dunno, you dunno what you're doing for who. Like, and that's what a lot of people do. They kind of like assume, I think. I think they assume that. Something [00:11:00] is needed. They build it, they ship the code, no one uses it. They select yes, we need more features. They add more features. Well that, that's not, that's not how it works.

[00:11:08] But anyway, one thing, but the few things from startup school that we, so yeah, and I, and I asked Maxim to join me and, and we started like remotely, it was beginning of 2020 there was some rumors about this big kind of flu stuff going on in China. And so we thought like, yeah, that, that's probably like the really.

[00:11:25] Which for people listening us from 2000 55 there was a thing called co covid. And so yeah, we, it thought like, okay, like the whole, the whole plan is gonna shut down, but that sounds like just the right time to quit my job and start, start new sketcher, any end. So we took the basics of Y Combin, which is like launch fast.

[00:11:47] Talk to your users, focus on everything. What's important. So we just said like, okay, we gonna give ourself two months to build some kind of MVP. We're gonna release it. And that's how it all, that's how it all [00:12:00] kind of started. In two months, we actually released our MVP. We actually, I wrote a blog post.

[00:12:05] There was like a blog post. How we've done it mostly on like. Credits from startups called and AWS and mostly using just AWS Lambdas and Dynamo. So it was like, and, and AWS Elasticsearch. So it was a lot of like cloud slash serverless solutions. So yeah, we just, we just kind of, we just kind of launched and surprisingly it was a really nice, it was a really nice thing to do because that got us some kind of like, attention from the crowd.

[00:12:33] I think I've done a lot of like passing on Reddit and hack news. To figure out like if people like it. So I think at this stage we kind of realized that yeah, like people like it. And we thought like, this is an, even though no one was really like, yeah, take my money and, and give me the access.

[00:12:49] But it felt like, yeah, that there's like some buzz around it. Like people, people think it's cool. So we should probably, we should probably keep doing that.

[00:12:57] Mike Bifulco: Sure. Yeah.

[00:12:58] You've taken a really [00:13:00] interesting path. So I, I would guess you may be the only person I've ever met who started coding by learning R and found your way into building a full on product. Like I think there's probably a lot of data scientists, data engineers, whatever they're called in 2024.

[00:13:12] Who, who learned R or Python or something like it first. It's definitely a fully inverted way to, to get yourself to learning about servers and the web and all that from there. And I'm, I am really intrigued by your ability to drum up interest too, right? Like, that's always an interesting part of the story for me as well.

[00:13:28] So you mentioned writing some blog posts and getting attention for your, your MVP and your early prototypes. It's the chicken and the egg problem, especially for something like what you've ended up building is really fascinating to me. And it sounds like you found the right, right ways to get people pointed to it.

[00:13:43] I, I saw earlier I was googling your name and saw your post about your trick to drum up. I think it's something like a thousand plus stars on your GitHub repos. Right. Things like

[00:13:52] Artem Bugara (NewsCatcher, CEO): yeah.

[00:13:53] Mike Bifulco: you, you've got a marketing muscle somewhere along the lines, and I'm, I'm curious how you. Landed on that?

[00:13:58] Like is that just a natural [00:14:00] thing for you? Is that something that you've gleaned from people you followed? How did that happen?

[00:14:04] Artem Bugara (NewsCatcher, CEO): Yeah. So I guess the, I, I'm trying to think. Yeah, so we have, I think I have in total like maybe 5K stars. I've done two like things, some really. Like it's, honestly, it's like a hundred lines of code in Python, but like done something useful, very connected, basically like the RSS feed of New York Times with just name New York Times and you type New York Times and you get all the latest news from our, it's RSS.

[00:14:25] It's something really easy, but no one thought about like doing, doing this, but I guess, yeah. So this, this kind of marketing or like, let's call it like go to market through developers is actually not, I, I don't think it's, it's right now it's not. It, it's, it's one of the, honestly, it's not what you should do in the beginning.

[00:14:47] What you should do is you should like, call people and like talk to them and like email them. But I probably started it 'cause I knew that like, you know, early on I. That's the only thing I actually, I, I, I, I think I thought I [00:15:00] understand how to write. Yeah. I was, I was writing some technical blog posts before, like in the last six years and stuff.

[00:15:05] So I thought, like, I feel like I understand really well how to write and how to talk to developers, and that's actually someone who I consider myself to be. So that was just the thing that I thought I will do the best, not. Maybe I also thought back then that this is like the right approach. So we were just more into like, you know, we'll just try, try something and let's say it's something that we probably assume that we know the best.

[00:15:30] So that's kind of why, I think that's why we decided to talk to like no, try to try do like I think it's, you can call it like the bottom up approach or like trying to talk to which, which actually ended up I think not being. I dunno. I think, you know, it's really hard to go and like write emails and talk to people and try to get some like decision makers to talk to you people with real problems.

[00:15:57] And I think, and it didn't help honestly, the fact that [00:16:00] we spoke just to developers. 'cause usually I. A lot of them, they themselves don't know what they do or what they need is like, you know, not really something critical. They just wanna play around with something. So I, I guess for some industry, maybe this works, but for our, that, that's, I don't think that's like kind of the, the right, the right approach.

[00:16:20] And I think like we lost a lot of time trying to talk to wrong people, basically.

[00:16:25] Mike Bifulco: Yeah.

[00:16:25] that is. Also an interesting angle too, and you, you've said this in a way that a lot of founders who have been through the process and have, have found their way into a product, say it, that in hindsight, there's definitely other things you could have done, but really the fundamentals that we hear a lot of, go talk to as many people as you can who you think might be your customers before you build something.

[00:16:44] Validate the idea, then spend time building it, and then, you know, iterate, keep talking to them over and over and things like that. I'm really interested too, that you came to this through YC Startup School and then from, from our sort of prior chats, I, I know you had an interesting way of even getting into your Y Combinator [00:17:00] batch too.

[00:17:00] Can you tell me a little bit about that?

[00:17:01] Artem Bugara (NewsCatcher, CEO): Yeah, so we applied to yc, I think first time We applied as soon as we started, and we applied in total of four times with new sketcher. And we got invited to interview twice with new sketcher, and we actually got accepted, got invited to YC twice. With new sketcher in summer 2021, we had to decline like going to Y Combinator because we already agreed to go to Tiny Seed, a small kind of program for Bootstrappers.

[00:17:31] So we took some money from them and took this like year to basically just like. Do our small bootstrap thing as, as we thought, like it doesn't have a big enough kind of market. And then basically we reapplied in 2022 for summer 2022. We got into interview again. It was with Michael sbo from 2021 to 2022.

[00:17:55] We grew from I think three. So we started, when we started we had like zero MRR. [00:18:00] Then when we applied in six months, in 2020, I think we had like, like few hundred bucks, maybe a seven bucks. MRR then we first accepted to, so it was our third application in 2021. We were at like three KMRR, but, and then, you know, they kind of like accepted us and we said no.

[00:18:18] I said, like, we explained it was all fine. I think big kudos to Brett Flora, who, who were the partner there. And we reapplied one year later and we kind of was like, kind of really, like, I spent on a application, probably like a day, really, like at this point I didn't like spend that much time there. But I think we, we got really like, you know, nervous because like right now we already got some money, so we had to show actually some growth.

[00:18:41] So we grew from three K to 20 K. And I guess, I don't know what, but like why C why C accepted us again and I was really happy we are like this Michael. 'cause I think. I was, I'm, I'm a big fan of I'm a big fan of his and I think few things for people who kind of wanna apply to yc. I know there's, I actually think I know why we got accepted, so I, I have some [00:19:00] like, kind of as assumption about why we got accepted. So a, a, a few things here. Like, first of all, we got really consistent, so we applied. First time we said we're gonna work on it. We applied Second time we said like, yeah, we actually worked on it.

[00:19:13] We applied this short time. We said like, yeah, you know, like, you didn't get back to us first two times, but look, we get some progress and so on and so on. I guess it's really important to show kind of, you know, the consistency that you, that you have. I guess a lot of people think like. That's what YC kind of actually assume.

[00:19:30] Like, okay, you have some idea, you say you want to commit to it, but like if we don't get, if we don't accept you, are you actually gonna work on it? 'cause like, usually I think you should, you should work on it. Whether you get to YC or not, YC shouldn't be. Or that's not the only way to start like a startup.

[00:19:47] And I guess, you know, I, I'm, I'm, I'm, last time there was like applied to YC on high canoes. There was like a lot of crowds saying how YC is. Bad or how a lot of companies do like millions [00:20:00] of a RR without funding, which is again, true. I, I don't say anything about it, but YC is trying to invest into people who like, say, like we are gonna build a billion dollar companies, or like trillion dollar companies, and obviously.

[00:20:14] That is a really different path from like, you know what you can do bootstrapping and you know, if you are solo founder and doing million dollars a year and spending on it, probably like three K bucks in a s bill. I mean, honestly that's, that's, that's absolutely fantastic. You just like cashing out all the money.

[00:20:31] But it's not a billion dollar company still, even though you can get paid more than founders that actually started a billion dollar companies. And one thing that stood out to me on the hack news, there was like, I think someone saying like, how. People in, in yc they like, you know, they need to go to Harvard or MIT or this or that.

[00:20:50] And, and I thought like, you know, it's, it, it's, it's probably true. It's true because the best kind of kids get accepted to M-I-T-M-I-T doesn't [00:21:00] really give them that much. Harvard, it's actually like they have such a high bar that obviously doesn't market for like kind of people, but YCI think has always been and still is a place for outsiders.

[00:21:10] And I guess the biggest kind of. The biggest biggest companies are built from outsiders, and I guess that's something you can probably tell, you can probably like say about us. Maxim and I are like first time founders. We never studied computer science. We got into coding. We started new schedule by quitting our jobs.

[00:21:30] We come from, we like kind of, we come from a country, from a region where the war started in 2000. 14. So we are like a typical exam and like we don't have any like cool kind of degree. The, the places we work at were like SMBs. We never worked at like Microsoft, apple. We had not like no good school, no good university, no good, like place of work.

[00:21:52] Like none of this is really like a requirement. So it's rather being consistent. And I think the last thing I'm gonna say about YC [00:22:00] and people get who try to get accepted to yc. I recommend a lot of people to YC and then I, they get to the interview and they, they get rejected and I ask them why, and they tell me like, yeah, they told us about because of this or that, and I'm looking at them and say like, look, don't worry.

[00:22:13] Like I don't have this figure out yet either. So that, that's like, that's one of the reason. But I guess the biggest one is like YC in people. They kind of, if you are like a great founder with like the, honestly, what you're gonna think is like the stupidest. Idea that never gonna work. They still gonna fund you.

[00:22:28] 'cause they know that like you are smart enough to like validate it, understand that it doesn't work. Change your mind and build something great. So YC really invest in people, not ideas or anything like that.

[00:22:41] Mike Bifulco: I concur and it's a hard thing to convey. I. You will see a lot of things online about people applying to YC and getting no message back, but getting rejected, not hearing back, getting an interview, and getting rejected. And I think the, the through line in your story, and from what I've seen and observed with my company and with others that have been through the program, is that [00:23:00] not only is consistency important and people being invested in the problem they're looking to solve, but showing some, penchant to listen to feedback and make change based on your business. So when they say they're, they're investing in founders, it is really like the type of person who wants to build a company that the company will be successful and not the type of person who has an idea. And it has to be that idea.

[00:23:20] Many YC companies have been through pretty wild, like 180 degree pivots that have moved them from, you know, one, one area of the world to something completely different. And, it's a subtle, weird thing. I like you. I'm also not an MIT or Stanford or Harvard or, you know, wherever Carnegie Mellon grad I went to a humble public state school here in the US and have stumbled through a lucky career at since then.

[00:23:40] But I, I, I think a lot of the folks who I met, especially going through, the summer program at yc at the very beginning, last summer, we kicked off with a an outing, a week long weekend where all of the YC founders for the batch went. And one of the things that was consistently true with every founder that I met was imposter syndrome.

[00:23:55] Everyone thought they didn't belong there for one reason or another. I'm too young, I'm too old. I haven't finished [00:24:00] college. I don't know if my idea is good. I'm not smart enough. All these things. That's, that's maybe hopefully a happy thing to hear for a lot of people. Like nobody really thinks they can do these things.

[00:24:09] And it's an important thing to realize that we're all fallible in some way and we're all you know, just, just trying our best. It's not like there's some secret handshake that.

[00:24:17] makes. YC founders better or different or whatever than everyone else? Yeah.

[00:24:21] I don't know. Maybe I should get off my high horse there too.

[00:24:23] There's definitely a lot of privilege wrapped up in what I'm saying, so yeah. Okay. So Artem you went through the YC program, that was your, your, you were summer of 2021 was your batch, so that was a few years ago now. Tell me about where news catcher is now.

[00:24:36] Artem Bugara (NewsCatcher, CEO): Yeah. So we in summer 2022, we went to yc.

[00:24:39] Mike Bifulco: 22. Sorry.

[00:24:40] Artem Bugara (NewsCatcher, CEO): yeah. So two years, yeah, two years ago. So so yeah, when we went to yc, we didn't, we had a lot of. Different types of customers, basically, like people doing probably like the, like a lot of platforms who need an integrated news or something like that.

[00:24:55] And I guess when I first meeting with Michael Sebel, he was like, okay guys, I like doing like, what is this? Like, is [00:25:00] it like for enterprises? And we said like, yeah, it's probably should be for enterprises. And he's like, okay, so what is your, like what are we trying to, like how many like calls and sales calls do you have?

[00:25:08] And I'm like, well, we, we don't. And he's like. Yeah, so you should probably go and find four and, and get four pilots before YCR. So you have pilots with real companies, with enterprises. And I kind of, I thought I, I did actually that, but then I realized that I have practically like four people interested enough, but not actually like, signed anything or like, not even like agreeing on the price.

[00:25:33] Well, long story short, it ended up really well we, we closed a few big contracts, I actually like, got into full y like I I, I flew to one guy to his like villa in Italy for like one day. And it was actually ended up being a lot of partying and then like next day, fortunately we, we got all him and I well, we both like woke up actually early, so we actually have like 20 minutes to discuss the business.

[00:25:58] We did some kind of like [00:26:00] handshake agreement that they're gonna become our customers and yeah, by demo day we basically like. Had like, nothing much changed, but we just had more customers and like more kind of idea that, that it should be working. So by our demo day, we already had like 400 450 KARR.

[00:26:18] Which, which isn't actually honestly that, that good. If anyone listen to me. It's really good if you have that after you, like just started yc. But if your company already exists for quite a while, that's also like, I don't know if, if that's a good result or not. But then basically we got we got great investors.

[00:26:34] Like we, I spent fundraising like two days actually, that's one of the magic of yc. And yeah, we got, we got, we, we got a small round that we just needed really fast and get back to work. And I think we kind of, I'd say, didn't do that. I think what we should have done, we, we didn't, we didn't do that. And so the 2023 was somewhat [00:27:00] like us being like adding more customers, but struggling to find like, where is the market, where is the product?

[00:27:08] Product, what is the problem? Where is the kind of feed? And I think only now. So we getting kind of hell of a year in 2024. I think. Partially because like largely, which models got like much more. A lot of people right now understand that like how easy it is to work with tax and now they need great sources of data.

[00:27:29] And one thing about us, we got really lucky 'cause we always doubled down on data quality while a lot of competitors like did on like NLP and stuff. And like with large language models, basically all UNLP is like now, like everyone is at the same level. You need to start from zero. And it got really easy.

[00:27:46] And I think we recently, like we are hopefully like I. Soon gonna end up on like, having almost already a hundred percent growth since the beginning of [00:28:00] 2024. And I think we, we, at the stage where I realized that what we should do is basically we should do like plenty of like deployments, more like onboardings.

[00:28:09] We have great data. We know that in this data, there is a lot of value for a lot of enterprise customers. We know that no one does those like bespoke kind of integration, well, especially big companies that can't afford it for the same price as we do. Even though we charge like six digits, like, you know, Microsoft engineer with like a million dollar salary.

[00:28:27] Like, you never gonna be able to like, you know, to, to afford the same stuff. And so basically what we do is like, we take our data we take a lot of like different LLM based like models. We take our customer's data, which is usually the, the trick is like, it's usually like different from, for each customer.

[00:28:44] And we say like, look, based on the data you have and the data that we have, we're gonna try to build like a, like a connection. Then we will try to make sure that like our data, like it was really work like. In the most like excellent way possible. And so you actually can [00:29:00] have, can have something that works really great and works really great for you.

[00:29:04] And the thing is like, and it's, and I don't see other way, like some of our competitors trying to build like kind of NLP and Reach news API and. What we hear from our customers, like, yeah, we try, we try this, but it doesn't work because like, yeah, it has a lot of different tools, but it's, it's never enough to like perfectly feed.

[00:29:23] And so we said like, like, look, just try to believe us. We have the more accurate, like the coolest data that you need. And we actually gonna spend a little bit of time making sure that like we like. To do the integration initially and then it's gonna work. And I guess that's our, that's our plan right now.

[00:29:40] And one of the beauty of this is like, it actually works really well because like, one, we can charge a lot for that. And like we can, we, you know, we can have, I. Like this onboarding, we can actually spend a lot of time and resources to make sure that it works, because again, we need to do it once and then it's like, you know, it, it, it, it just working, doing the same [00:30:00] thing with obviously doing some kind of like supporting and maintenance in between, but it works.

[00:30:05] And two, that's like, I think like the only, like the big superpower that we have is that we can invest all those like. Custom thing and like a lot of like our time to actually make those connections. And that's one of the things that like the big companies can do. And that's one of the very few, that's what YC says.

[00:30:27] Like you need to do. Like really, you need to make your customers really, really, really happy because you can and no one else can. And one of the best thing is like. It makes much more sense when you do like big checks because if you build in something that like, technically you'll have to scale to like, let's say like million of users you know, custom, you still need to like do a lot of like talk to your customers e each and every one.

[00:30:50] But it's, it's really hard economically to do for really long when they pay you five bucks or like 50 bucks or 500 bucks and it's much easier [00:31:00] when they pay you half. You can, you can extend this like kind of, you can extend this period of time for much, much longer if you, if your customers pay you much more and like continuously learn and any end, we are at this stage where we're like trying to build a lot of those different connectors.

[00:31:15] For a lot of different use cases, but at some point we, like, it's know every additional customer get like, it, it, it becomes easier and easier for us to do so. So hopefully at some point, which is actually like I, I do this, I just recently started to learn about like Palantir and that's what, like Palantir have been doing this for like 10 plus years and only recently they started doing actually products that kind of like somewhat scalable.

[00:31:38] So I think Palantir is like a great, great example of like how. Given those bespoke integration, actually can, can still make you like, I don't know, like a, a multi-billion dollar company.

[00:31:53] Mike Bifulco: It is a the economic match there is really interesting. Like the Palantir is a, is a quiet giant I think [00:32:00] for a lot of people. Like I, I, I think possibly the audience of this podcast is, is more familiar with Palantir than let's say 95% of the world. But the way that businesses like that grow and become successful has a lot to do with their early plans and like the targets that they, they try and land and.

[00:32:15] Getting, getting your business to a state where you know, you're growing by leaps and bounds is also something that requires that level of strategy. And as what I would imagine is, is a, a scrappy little company in comparison to Palantir. There's different ways to do that. Definitely a different ends to those means.

[00:32:28] But it's certainly possible too. And, and I think it's a fantastic story and a really cool thing to see. Like, just, just how you've grown from like you know, a project in your flat and, and suffering through learning r and moving to Dubai and all these things. To something that's really a, a growing and inertia gaining project over the years and especially now, I would imagine being in a season where news is going to be more and more important.

[00:32:50] And LLM certainly have gained an impossible amount of popularity. It's probably an interesting year for you is, is so. I should couch this in the fact that you and I are recording this the [00:33:00] morning after the very first US presidential debate which I could not watch last night. I just didn't have it in me to turn it on for, for a million different reasons.

[00:33:07] But do you, do you imagine that the sort of political the, the election season in the United States will ratchet up user base for you? Do you think that'll have an impact on your end users?

[00:33:17] Artem Bugara (NewsCatcher, CEO): I think not really. And the reason for that is like, it's exactly what I was talking about. Like our customers, like, I think such thing as like. Like, you know, the presidential debate is, is usually something that like stakeholders, like chiefs should just read and understand. There's like nothing that much that you should do.

[00:33:39] However, if. A company that owes your company money, has some small news that their branch getting closed, that they like, kind of had a fire or something, or that they actually like a five news local news this year that they are changing. The c or that's what you kind of, that's what is like a, like a, [00:34:00] like a stuff that people really don't think about, but then they, but one, but one day it like hits back on them and.

[00:34:06] And then they come and see like, okay, look, every time someone doesn't pay us, we actually, so where one bank says like, they, they don't pay us. And we go and check like why this happened, how could we know this

[00:34:16] Mike Bifulco: Hmm.

[00:34:17] Artem Bugara (NewsCatcher, CEO): And they go, we go and check the news. And we realize that like, okay, probably if we were reading like the local newspapers, we would knew that months ahead that probably they not doing right.

[00:34:26] And like, and sometimes, sometimes they're like, and we even give them more money when they ask, but like financial reports were like, fine, but. Financial reports are not something you get on like every day and like, you know, probably if, like, yeah, if, if if company changes management like really frequently, maybe there is like, something wrong with it or, or like, all of those facts.

[00:34:48] So like all of those things like, it, it can like, you know. It's not probably like going to be like a billion dollar change for each company, but there is a lot of those small, [00:35:00] small little things that can be, that can be captured and like you honestly like missing those or. Directly, or like indirectly can, can cost like tens of millions of dollars for, for per department of big companies.

[00:35:15] And that's kind of where we're trying to, where we're trying, but obviously like there's always gonna be like people trying to build like news, website coverage, comparisons, sentiment or, or Trump and Biden. And it never, and it never kind of like, it never monetizes into anything I feel like. But yeah, so we are like trying to do a little bit more like.

[00:35:33] Not mainstream stuff. And that's why I say like, you know, some companies think that they follow news while invalidate, they follow like the biggest news that everyone follows, but they don't really follow news that can affect their business in a day to day.

[00:35:48] Mike Bifulco: Of course. And for you, having, having this strategy must have a leveling effect where seasonality is not something that you have to deal with quite as much. And from, from a founder's perspective, it's nice to have predictability and growth [00:36:00] and stability maybe in, in ups and

[00:36:02] Artem Bugara (NewsCatcher, CEO): Yeah. Yeah. And that's, that's, that's another really nice thing about like enterprises and like doing deep integration. 'cause like, honestly, if we succeed and if it works, I don't know what has to happen for, for them to like go to someone else or like to do something else. 'cause like, even like, because then even just to change it to someone, it is like already gonna be like a.

[00:36:24] Like maybe a half a million investment for company, like in terms of like salaries and change and all. They, they will like customers, if you do something that fits like a glove and it works for them, customers, like they're gonna stick with you. Like your LTV is gonna be like huge because like, unless they, unless it doesn't bring them real value or unless they kind of run out of business, run out of money, you are kind of like you are married forever.

[00:36:49] Mike Bifulco: Yeah, no doubt. Yeah. Okay. So let's, let's jump into the nitty gritty of it. Tell me about the, your, your tech stack. Like the API developers are, are thirsty for nerdy [00:37:00] details at this point. So tell me what how you've built news catcher in the the bits and pieces you've tied together to make it all work.

[00:37:06] Artem Bugara (NewsCatcher, CEO): Okay. That's funny. So long story short, we have everything is like. Deployed in the, on-prem on Kubernetes. Mostly. One thing I wanna say is like, it's not something really you should worry about is, especially when you're like a small startup, you, you have to start as fast as you can. And that's why we started with AWS and Lambda function.

[00:37:32] But then one thing we found out is like, if we get, like, if for, for whatever reason, you know, we are like a small customer to or GCP or anyone else, if we get. Kicked out of them. Not only are we gonna have a problem that like we just need to find other servers, but also gonna get bounded to like, like proprietary like services.

[00:37:54] And it's sometimes really hard to like change those. So we said like, look, we're gonna do everything open source, everything's gonna be [00:38:00] infrastructure as a code. And you know, if something happens or like we need to scale, we can just like deploy the same kind of stuff on other clusters, on other, like on other hardware.

[00:38:09] And that's what we do. So it's all Kubernetes and like a lot of dockers a lot of different microservices, I assume that like, are connected together. And our main stack is like Python MQ for like event processing. Elasticsearch for indexing the data. And to be fairly honest, there's probably like 50 other different technologies, but they like.

[00:38:29] Small

[00:38:30] Mike Bifulco: That makes sense.

[00:38:31] Artem Bugara (NewsCatcher, CEO): Like fast API and, and we do, you know, like a lot of right now, like basic LLM, we've like fine tuned, we fine tune like mis trial seven B for example. Actually one, one thing I would, I would, okay, we know their customers. I think we're using their open source service, but I think they're really great company.

[00:38:48] Open, open pipe by. It's a great company for help you fine tune

[00:38:56] Mike Bifulco: cool. Open

[00:38:57] Artem Bugara (NewsCatcher, CEO): yeah. Open pipe. Yeah. Yeah. I think [00:39:00] the guy, the founder, he's actually an ex y Combinator. He's an ex XY Combinator software engineer. Yeah. Oh, he has a, basically a great lending. He has a great lending page. It, it has some great ROI numbers, 25 times cheaper than Gvd four.

[00:39:14] 5 million to start and $17 million saved by our customers this year. That's a great, great, great.

[00:39:20] Mike Bifulco: That's a solid pitch.

[00:39:22] Artem Bugara (NewsCatcher, CEO): So to be That's what I like. I, I like people like trying to understand the value. Yeah. So we use an open pipe, for example, for, for fine tuning some stuff and yeah, it's, it's all kind of open source. No, no kind of proprietary not much proprietary stuff.

[00:39:39] Mike Bifulco: Okay. And tell me about news catcher's story from a open source perspective. Have you, you made any of your software available for people to peek at?

[00:39:48] Artem Bugara (NewsCatcher, CEO): Yeah, we don't really have other software available just because it's not really kind of possible. But we have two things. I have a few open source project. One is actually called New Sketcher, the one that we done while we are just like doing our [00:40:00] MVP, which kind of gives you, and I don't maintain either of them by the way.

[00:40:04] So News Sketch and I one py, Google News, like is basically like a, like RSS feed documentation of Google News, which I had to like hack around myself 'cause there was like no documentation. So basically Google News has r ss feed and it doesn't have documentation. There's like a lot of parameters that I had to like, kind of deduct and try to like, just hard, like try one after another one to see what is actually exists, what works, what doesn't work.

[00:40:28] And another thing, we actually have free access for researchers. We have a bunch of research papers. I think the most interesting is one of the recent one by Berkeley where they see how well LLM can forecast the future using like news. And they use our, our API there. So that's like an interesting one.

[00:40:49] So yeah, if you are like a researcher, if you're a student and you wanna mess around with like news and that's going to be your research, just go to us. We have free access for [00:41:00] researchers and we are really proud that we, that we do that and like, yeah, we, for example, also help to big nonprofit organizations, NGOs, like we should, I dunno, like, so we sign that we have like, we're gonna help Transparency International, catching all the news coverage about corruption and briberies and stuff like that.

[00:41:22] So there's like a lot of cool stuff we do for free.

[00:41:25] Mike Bifulco: I love that. I think it's really an important thing to have some access to folks who can benefit from your product, who never could afford it, and that it speaks a lot to what you've built and certainly the value that you've provided for your customers that you can, you know, in some way subsidize people who are doing research and things like that.

[00:41:40] Artem, I'll make sure to, to reach out to you and grab links to the research papers that you mentioned. 'cause I think that'd be something that'd be one fascinating to read through and I'm sure listeners would like to get to. But I think it may also give some indication for some of the folks who listen the sorts of projects that might be interesting for people to pursue if they're looking at news catcher or if they might be a good fit as a user of yours.

[00:41:59] [00:42:00] I'm curious then. For the folks in the audience listening, how would they know that news catcher might be something that they should explore as a use case for their company?

[00:42:08] Artem Bugara (NewsCatcher, CEO): Gosh, that's a, that's a great, that's a great thing to, that we actually don't know answer yet really well. But, so globally, what I say is like is, is every industry has really different things like, some industries have to like do adverse media checks, for example. You know, like you open a bank account, they probably need to run like an adverse, like check that.

[00:42:29] Like there's no news about you doing like, I don't know what, but, and there are some companies that just like can benefit and sometimes they just like don't know what they like, it, it just like, they use too much manual time for that. So what usually I try to say to everyone is like, look, if there is, like if you can.

[00:42:48] If there are any news that are happening either locally or globally or in other countries, whatever it is on your industry that you, you know, like your business can directly benefit [00:43:00] from. For example, I don't know, like we have companies who like sales intelligence, they say like, like, look, we just need to know where, when our customers raise more money, open new business or something like that, and they say, we do it manually.

[00:43:13] We lose a lot of time on this. As long as you can like articulate clearly what kind of news you have and like, as long as you can say like what kind of answers you need from each news article, we basically gonna make sure that you kind of receive those at the right kind of, at the right time, at the right place.

[00:43:31] And a lot of people just like don't know how. A lot of stuff is possible. Like even one, one of my latest discoveries, like, like financial like, or like where people invest money. We have customer right now with like private equity in real estate and they wanna know if there's like any. Transport, public transportation infrastructure expansion going like, if, you know, like, if there's like a bus station opening somewhere, like the prices of the multifamily apartments gonna go up there.

[00:43:56] So they want to know that. And there's like hundreds of those news, like every day, [00:44:00] for example, or like a financial institution. They not always invest in Apple and Microsoft or Palantir. Sometimes they buy bonds from like. Hospitals and universities and schools, and they kind of need to know local news about the schools, like if everything is happening there, if they're gonna fail on this bond.

[00:44:20] And there's like a lot of those stuff where kind of, I guess people know that like there are news and they just like. Don't know how to scale it or they just never thought that they could do. And so we kind of try to help those that like figure out that there was like something in the news that is like important to their business and they come to us and we try to like make sure that like we integrate it like into their existing workforce, into the systems perfectly.

[00:44:44] Mike Bifulco: .

[00:44:44] The summary I'm hearing perhaps is that if you, if your business or your strategy is swayed by. Signals in the news and, and you want to trigger some sort of reporting or workflows or decision making or whatever it may be. This is a good use case for plugging in the news catcher and trying to [00:45:00] listen to the, the broad swath of places where news can come

[00:45:03] Artem Bugara (NewsCatcher, CEO): Yeah, and I, and I would say, I would say like the only thing is like, it doesn't have to be some, like we work with some like Foresight Trends team. It can be that, but it's usually like much, much more. Specific, but which is still for some kind of, you know, companies is like, you know, like a hundred million dollars line of business that gets like affected by small news every day and they just like do, don't do anything about that.

[00:45:24] And so just by changing this, just like a slightest beat can, can save you millions of dollars every year.

[00:45:30] Mike Bifulco: Well, Artem thank you so much for joining. Before I let you go, I have a couple of important questions for you. First of all, are you hiring, are you guys looking for anyone at the moment?

[00:45:38] Artem Bugara (NewsCatcher, CEO): Yeah, we kind of always hiring right now. So we are looking for, I think right now we're gonna look for two positions. We just hired a first technical like kind of account executives and we're gonna probably look for one more, so I. What I'm trying to do, like we do, like I said, a lot of like bespoke integrations.

[00:45:54] So you need to be really, really technical. You need to, you're not gonna have a sales engineer [00:46:00] with you. You are actually gonna have to understand like how things work. You need to understand like customer problems and try to like come up with some smart solutions and things. So we are looking for technical account executive, AKA, the closer, the guy who, who actually does the sale.

[00:46:17] And then we probably gonna look for very first, like LLM AI engineer slash data scientist, I dunno how you call them today. We do a lot of like, so basically we do a bunch of like fine tuning and stuff. We do really basic things. Surprisingly they work and probably right now it's the time for us to take it more seriously.

[00:46:35] So, I don't know if you are passionate about AI and LLMs, we know, I cannot promise you like a million dollar Nvidia cluster. , actually, I can promise we're not gonna have that, but we can deliver something like zero to one with just. Little resources, but , measure the output, and just, you know, we have a lot of, you know, one thing I learned is like working with [00:47:00] enterprises, really few of them actually need this, like a hundred millisecond latency stuff.

[00:47:04] Some of them can wait day or like 15 minutes. You can do a lot of right now with LMS, if you have like 15 minutes windows. So yeah, we are looking for like, for a great like AI engineer. We don't have those posts together. Probably by the time this podcast is live. You can, you can kind of apply.

[00:47:21] Yeah. We are looking, we are looking for, and yeah, we mostly, I'm sorry for folks in the US who I think like the majority of the crowd, but we must likely need someone to work around the European time zone as we are all in, in Europe. Yeah.

[00:47:35] Mike Bifulco: .

[00:47:35] Amazing. Artem, of course. , I'll throw those in the links to any job openings you have in the description here. The other thing I wanted to ask, and I don't actually think we've said it out loud on the podcast yet news Catcher, what's your website?

[00:47:45] Artem Bugara (NewsCatcher, CEO): Yeah, it's news catcher api.com.

[00:47:50] Mike Bifulco: Love it. And where's the best place for people to find you if they're looking to chat with you? Artem.

[00:47:55] Artem Bugara (NewsCatcher, CEO): Yeah. If you are ever in Warsaw, Poland. Please, [00:48:00] like really reach out to me. I don't have, like, it's not, not that many people here I know. So I'd be happy to ask you out for a coffee or something.

[00:48:09] Mike Bifulco: Cool. , that's amazing. I would hope that people take you up on that. I've heard Warsaw's an amazing place too. I'd, I'd love to visit sometime. Artem, thanks so much for joining, . I really appreciate it. It's been fantastic chatting with you and , what an interesting product journey you've had.

[00:48:22] Anytime you're you're interested in coming back on the show, we'd love to have you. Thanks so much for joining.

[00:48:27] Artem Bugara (NewsCatcher, CEO): Thanks a lot. Thanks

[00:48:28] Mike Bifulco: All right. Take care.