Open Core, Pricing and AI product development with Dat Tran

Hi, everyone. My name is Patrick Akio and if you're interested in AI product development pricing or open source and open core, this episode is for you. Joining me today, is that tran? Who's the co-founder and CTO of price Loop? Do AI aperol his links to socials in the description below. Check him out. And with that being said, enjoy the episode. Have you played around with chat

GPT? That's amazing stuff. Yeah yes I did. So I did a few coding related questions. So for example, write a sequel, where your filter for days or also like python chilling rated things. For example write me a linear regression so that it solves this and that yeah in. Yeah, the answers are Right, nice. And but, yeah, I mean those are very simple questions, right? I don't think that, you know, if you would ask a very hard question that, you know, chechi would come up with, you know,

like a proper answer. So, like, I mean, for simple question, I think it's very good. It's better than when you would do it on Google or also most likely on SEC overflow. And also I would even say I don't even Simple question on stack Overflow, right?

So for me it's like really when I would go to stack Overflow, these are more like very challenging questions that people haven't asked before, but I think that chechi could, you know, be kind of the first point of contact for those more simple questions. Yeah, I think it's a really good Baseline and like a tool to have in your toolbox at least that's what it seems to be like on the surface level.

I've asked questions here and there and it very confidently said, okay this is the answer and then I put it in doesn't even compile. And then like yeah that's not the answer is mistakes here and there but it to me it's very interesting like it's a first building block on a future which is kind of unknown at least to me. And it seems. Yeah I'm Gonna Keep a close eye to it. I'll just put it like that I

mean I mean at the end of day it's not really like super crazy a I you know that is doing something out of Box, right? So at the end of the day Shetty Beauty was trained on data, right? So especially in this case, conversion conversational data that people ask before in the past. So if you probably would look for like something that is out of the box, that no one has asked before, right? Or especially when there's a new

technology, right? And you would ask this to shake your booty, You will probably get that nothing but what it does well. And I think this is a cool thing in all this generative. AI really topics you get a since synthesis of this, right? So you get a mixture of different components and I would definitely see like automation automation related topics for that. So there's definitely a case for this kind of models. Yeah, that makes a lot of sense that it knows what is already

kind of tried and true. Proven based on the history that we have, but future topics. And when your trailblazing is not going to help you because you're doing in your investigating and you're experimenting and it probably can do those based on assumptions. But so can use of like your answers May Vary there, that makes a lot of sense. But to me, I've been involved in

a lot of product development, but not necessarily with an AI component to it. And I looked a little bit into your past and you've put a lot of products in production with an AI component to it. Like, is that different from like Product development, without the AI component. What makes it interesting putting that into production? I would say it's not very completely different, right? So, it is still a product development and it's the same

things apply there, right? Yeah. So you're still, you know, test for for an MVP, you need to think about the user interface and also the user experience

that comes along with that. But I think what's more Like the difference between like classical standard products, you know, without she know top of this is actually the user feedback that comes to it to it. I mean, obviously in traditional proactive element, you know, if you create a website Standalone website your ass, okay is does a user gets it, you know, is it simple enough? So that you use, it doesn't get any guidance to use the product, right?

Yeah. In machine learning, obviously there's this stronger feedback component because you need To collect data, right? Or like the output of the of what you created, so that you feed the model again, right? So that the models improve at the end of day and this is cycle is an iterative cycle. You know that basically hey you get kind of feedback from the model and then you know you can reuse it again. This is very important. For example, there are many cases where where you need.

It recommendation engines, right? So recommendation engines like, Obviously the most stupid or the most simple way to do in recognition engine is, I mean, I don't, I wouldn't call AI but you just take the most freegan clicks you know, on the product you show it or oh, hey are you basically just show like like a list that you create yourself,

right? But then you know you need to basically run, you know, more of this recommendation and like sections and the product and to collect this data so that you can create Eight more fancier models, right? And even though Spencer models, they only reinforce basically what the users like. So you will never, you know, get like like an out-of-the-box recommendation.

So if you have new products, you know, that you want to introduce, you know, you need a better way how to balance, basically, the algorithm are rhythmic part of the product and also, you know, the new part of this. So there's a lot of things that you need to think about that. And also, what I also see Especially Machining into products. I usually never see machine learning as a standalone feature or product itself, for example, even like, just take GPT, I think it's nice, right?

Like just to try it out. I mean, I think I read that it took Jesse PT 3 days to get, like 1 million users, right? Which is crazy. Just which is quite amazing. But obviously what you pay for it which, you know, what is the use of it, right? It's a fun. It's a fun project and you know it's really nice to try it out but usually machine learning is part of overall product, right? So you really need to think together with design and also with product management in general, right?

How does it fit as she overall into the product? I like that because that's kind of demystifying it as a whole. Because if I hear machine learning and if I hurry, I like I've read a lot of things and I've talked to a lot of people that are more knowledgeable about that but before Doing. So I was like, man, that sounds pretty complex, but if you put it like that, like, it is a component of your product and the whole product development life cycle in and of its own

doesn't really change. You just have to accommodate for this component and that can be gathering the user feedback and feeding it back into the model, to refine it, and to hone it, and to make it better. Then, yeah, it's just a component of a product that you're building it. And for chat upt, it can be a component of a product but it is

not like it is right now. Right now we just have the Onan and we can play around with it. But like you said, I wouldn't pay for it maybe for like a month to try out. But yeah, I've already stopped trying it out because I was like, oh, this is fun. And then that was that. Yeah. But when we zoom in to what

you're doing now, I saw your work very much involved in price Lupe. I like you're one of the co-founders in your CTO in there, as well. Can you explain to us what problem does it solve to start off with? Yeah so at Presto we actually physically bullying New another way of how business can run the pricing. And if you, if you, I mean, I'm not, I'm not quite aware of how much you are into the pricing space. But today's pricing is very

manual or rule base, right? And especially when we think about like our first customer segments which is Amazon. So there are a lot of Amazon sellers who basically, you know, they put the prices. Once you know, when they started the shop and obviously they do promotions but nothing else, right? So what we are, what we are aiming to do is is for them to create a solution so that they can make better data-driven decisions, right?

Because because prices they change because especially in times like that, where we are, you know, going into like a recession or we are already in the recession. We're basically inflation Drive costs vary up, and you need to As you increase your price. Because if you stay at the same price, now right now, the same time, you're basically making a

loss, right? And what we also billing, or what we also realize is that pricing, as I said is manual and full pace and it will still be rule-based and manual for many, many, many companies, it's very complex.

So we actually, you know, have customers who obviously have thousands of products and within this products, you have Different marketplaces, right to have Germany, you have friends, you have, you know, whatever marketplaces situation can exit and then you are also variations out of this product.

So you have a main product, you have radiation, A variation, B version C. And then you have your complex rules that says, okay variation, A variation B needs to be priced like that variation C variation D, needs to be pressed like that. So it's a very like, kind of like a complex like if then block, I Say like this right interest and therefore what we built is we built a also, no code spreadsheet which actually handles like large data months.

And it's the UI is as I said, still resembles like a spreadsheet because many of our users especially pricing managers, they, you know, they work before with with, with Excel and and also Google Sheets. So they are actually used to that kind of like user interface before. And I always found the idea actually quite interesting, you know, that to take kind of like a spreadsheet in Excel as a UI because in my in my previous jobs, when I was a TV, ah, Louis.

So it's a, it's a price comparison website here in Germany. It's actually one of the Europe's biggest price comparison service. I actually have, for example, like the CM team to set up kind of like, Automated ready to process with data because they want to wanted to do data-driven campaigns. That was taking basically back-end data from from our,

from our service, right? They should though was that, you know, as you know, right in an IT company, especially in Austin, like maybe like a normal company, you have so much priorities, right? So business as hey, can you please do this for us? I, we created this already. The idea, you know, on the Whiteboard. We also, you know, have a have a small proof of concept in Excel, but how do we automate that, you know, how do we basically get this into, like this time into Salesforce?

Because this is what the CM application that the team use. And, and yeah, you know, I was like, yeah, you know, like the main, I teach him, didn't really have the time for that. So I basically built a proof of proof of concept, you know. And that also It into production. So at that time I used Apache, Apache initially is a kind of a work flow pipeline tool. It's very old-fashioned, but the idea is, is really nice, you know. So you you have a GUI, you can basically put like a sauce in

that like a data source. You can basically do some transformation and then you can set it to 2 to the domain system, right? And that's basically what we also during right now with About this new code spreadsheet, tool, you can get data into our solution. You can basically do data transformation in a very simple way. It's a pipeline, right? Because every column, if you think about it, it is a flow,

right? Because because the column can depend on another column and we, and basically, you can put up your rules and then, you can take that basically into your production system again, because behind that we also put a lot of sub engineering best practices into into the solutions. So for example, It's typesafe. Which means that basically the type that we put into a column, you can only be either be a number be a string, you know, also be Jason.

So we introduced this Json data type as well because when you think about Excel and Google spreadsheets you can have any type but you can put your you can do it. Yeah. You can put a number there and then you change, you know, you put a string. And if you would take that sheet into your production system, right? I can tell you it will break Take in a production system or also.

Another thing is obviously it's not a database, right Excel or Google Sheets, which means, for example, if you have like thousands of data, so for now, we have one customer. We actually did a proof of concept with Google sheet and the amount of data actually increased to just 300 thousand rows, I think with 20 columns. And you know, when I would search for for value now, also just Go through the sheet, it's really laggy. So it's really really slow

already. So and yeah we're trying to take in all the disadvantages of you know using excellence and Google Google sheet into our tool and become more like you know the better middleware between basically business and also the IT department very interesting. I thought because I've always

seen people from or business since work with spreadsheets and I've always thought like don't we have a better way of visualizing the data. So it's interesting to me that you still I mean you remove kind of the pitfalls of the previous spreadsheets that are there and you improve on those but the visualization is still what people are familiar with because it's still a spreadsheet and at the end of the day have you considered like different visualizations or try different

things in there as well? Yes. Actually this is actually a very good question because like when I when I started price soup together with my co-founder at a time I always thought we're going to take that flow sharp like flow base basically pipeline as our main basically user interface. I mean there's there's many many tools out there that uses already right? So you have a prefect airflow right where they Have this this photo shot obviously it's no, not no Code and low code

completely, right? So you really have to write a lot of code to great as pipelines, but I always when I, when I, when I, when I thought about that, I thought that was. But a lot of limitations that you have with this flow charts, I mean, it's very easy for people to see flows, right? You have a sauce, you have a saint, you know, you said a note like, from from A, to B, right? Or to something else.

And then you see the flow Of you know, things going through but it's a bit in transparent but because if you work with data and such a lot of data, you do transformation in between these flows but you really don't see, you know how the transformation,

you know, in the data. So and it's actually quite costly because I always thought you know, you take every node and then you do real transformation on every node and then you end up something you know with like maybe phrase I will find the optimal price or final optimal output. Yeah. In between, when I see the flow. It's very interesting parents, It's Magic.

So yeah, it's magic, right? And it, I thought about Excel or who sheet and because at that time we were doing our first MVP, our first entry was completely different from what the bill today. It was like, you know, this is standard dashboard. I mean, a lot of a lot of

solutions use dashboard, right? So if you if you think about like, Like like typical SAS applications, they do like a normal dashboard, you know, you go in, you have your functionalities to automate some stuff you see numbers in sun, it's pretty nice, but it's very inflexible, right? So that, that is something that we started with.

But we knew that we need something more flexible because every of a customer have different rules, different logic, you know, they want to do different things, different Industries, right? So at the moment, we serve the e-commerce. Industry.

But you know, when you go to a license, when you go to a hotel, they have much different different kind of rules and logic there and then and then over the time I realized that like my co-founder and also one of even one of our data engineer and also our project Venture they use Excel and especially in this case, Google sheet to solve the problems before we actually turn it into SQL code and other python, right? So I was like, wow, why, why can we not?

You know, just take, like, the solution that we had already in Google sheet, right? And put this into production, it's tough instead of actually translating, you know, the the solution that we put in an extra sheet into SQL or python again, right? Yeah, so I was really like em, okay, and then I thought, oh wow, yeah, an extra sheet ingush, it is actually also just a flow but it's more visible. It's more transparent because Every column there we do.

You can see actually the transformation already, right? And you can see the flow of the data. And the coolest thing is with our to read. Now, we actually like in the core, we actually built a dependency craft. So it's, um, it's a feature that we have also on our roadmap, that actually, when you would, for example, have different tables, you know, different sheets that we would call it and then you could add like 20-30 columns, That kind of depends on each other but also in different

tables. And then if you would click onto one of the column, you can actually see the dependency, you know, Focus from from the source to the to, basically, this final column. So, this is something that, you know, I haven't seen in any other like spreadsheet tool so far and it's very actually important, you know, for for for especially businesses because they they create Logics, right?

And they want to see. Okay, what is actually, you know, the higher logic Based on all the tables and other columns that I created so far. Yeah, I love that because in my head exactly as you mentioned I'd like a flow chart or it depends because if you have a lot of data, it's it. That's also gonna take a lot of like overhead and like, cognitive capacity. So I like that. You iterate on top of that. You came in with an assumption,

you were like, okay, we can improve on this. And then at the end of the day, your customers were still like, working with spreadsheets and working around what you saw. So you took that with you and you're like, okay then apparently this is A good Baseline but we can still improve on top of this and probably if you look at the end product, now it's completely different that we've started off as an MVP. But that is exactly what you

need. At the end of the day, to make your product succeed, listening to your customers, and actually letting go of your assumptions that you initially thought were right, right. Even though that is completely going to bias you, you also have to let go of that. If it turns out to be wrong due to practice, basically, that's really good. Yeah, definitely. I mean, I mean, this is always Like the thing, right? So you start with that possesses right? You start with our own ideas. Yeah.

And then you create basically like like a mock, so that's actually what we do. So we actually create box and then, you know, we also talked to customers, we do user research also in sales pitches, right? We're also trying to get a gut feeling of okay, that's a does the user get it right because he understands the value out of this and then you should like Eight a proof of concept. Right? And then really understands

okay, does it make sense? So, for example, when, when like, after our first MVP, so I was really unsure if basically the table, right? The table is you. I will actually be like understood kind of by pressing Ventures and so on or in general by by our users. And then you know, we basically first like Creator or next MVP on top of the table and yeah, basically the Like to, to the this one customer that had and if you clearly understood already how to use it, right? Because it was a table.

Yeah. So but obviously, in this table that we did, it was a data table. So it was not like an Excel spreadsheet that you can do this. So, what we put there was basically yeah, you can turn of, you know, like a button like a toggle, right? You can change basically some values but it's not really like you can go in and create a column and do this and that it was already. Ready. Flexible enough, right?

So at the time, so you, we had a back end where we basically, we put some of our like blocks that we caught. So for example, Like A Min matching, block min, max price block. And basically, we could for every new customer, we could basically, you know, like create like a Json file that we had in our back end and then basically deployed for every customer differently, but the problem was at the time. Wow. If we would need to do that for Every customer, right?

Every customer to create those Json files, that would actually, you know, be very, very unscalable it but it was fine because we created this MVP in a fast. Way to get an understanding of, does it table word and does a customer gets it and and the feedback was really amazing. So, you know, he, he understood it, you know, he knew how to use it and also had ideas. Oh wow.

So if I can, I probably add another, you know, like Like another block to it. So another rule block and he said, yeah, you can do it but you need code. So you need us to do that for you, right? But but the feedback was actually, wow. It would be actually quite cool if he could do it.

Right? Is the person you know can create this book by themselves by himself, you know, based on documentation and and therefore you know, the idea of this flexible nurko tool was, you know, then then exactly one in a way. Because then We got all the requirements together, like, you know, how we should structure that. And yeah, now you can actually try it out so you can go to, to price to up. You know, we have our ASL version release, so release.

It like one and a half months ago and yeah, we're developing really hard. I mean, it's, I can tell you it's a very complex product to build on. I could imagine we just based on the things you explain like pricing and competitiveness there like with Variance and different marketplaces and different countries, even in different competitors. Like that's already a lot of data especially, I mean, I have a little history in e-commerce. So I know like the various amounts of products that can go

in there plus the variance also. Yeah that I can see how that gains in complexity there but you me talk before the show and

you mentioned something interesting in that you're thinking about or you already have open source part of your code or even all of it. Can you walk me through that thought process like why, why do you Think, or are you thinking of open-sourcing a lot of things? Yeah. So so right now we're only open source part of it. So which means that, for example, how you can create custom functions, also, basically our API connection with python.

So to our to our, our platform. And also we are basically, you know, cleaning up our our entire project. Because, you know, when you Help. So we have been developed obviously in South North 44 for a couple of months, right? You had a lot of kind of technical debt. I would call it, you know, a lot of other clean code and son and we need to clean it up before we gonna you know, Go full blown open. But yeah, in terms of your question like what is the

thought process around that? So I'm not quite sure if you like familiar with the term modern data stack. Yes, please. The modern data stack is basically like an open source ecosystem which comprises of basically tools that modern data people would use, right? And if we you know like just think about it again like what is our tool or what is what is the what is the tool that we created during? It's also a data tool, right?

Because people move data into our solution and then you know, the They do some transformation with that with our UI. And then, you know, they can export it back to any of their production system. Yeah, maybe it's an Erp, maybe it's a database right, and within this modern data stack, you know, you have different tools that. So you have a bite for ETL, elt you have prefect and outro for example, for orchestration, you have DBT, right?

So for basically, Of a modern sequel transformation and versioning, right? So everything around transformation and then and then you have also like dashboard Solutions, like light Dash and so on, right? So that basically combines idea of DPT and basically adding a nice bi Solution on top of this. Yeah. And at the end of day, when I, when I, when I think about pricing in general, pricing is a data project, right? So in pricing you, you Take data, so your take different

data sources. Obviously internal data sources from Erp systems, but they also, you combine it basically with external sources, right? So compare of tracking data, right? Maybe weather data and many other data that you can, you would need for your, for your, for your pricing. In general, right? Yeah, then you would clean this data, you would combine this data together, do some joists, and then, you know, maybe you have like rule base Logic on top of this or also like a machine.

Base model and then you export that data, right? So within this whole ecosystem, you probably use some of these modern data stack, right? Like, like are by to move data from HP, use db2 to transform the data. And then maybe use light Dash or a or superset, right? One of the like visualization tools to visualize the data in our tool is also in between of the modern data stack because at the end of the day, you know, we We want to connect with with,

with our bike, right? So we want to know at some point, basically having our bike connected to our solution so that, you know, people move data to our solution to our platform and then they, you know, can can do their custom logic in there then transformation. We also offer like a like a way for developers to write their own custom functions.

So you can for example write your function in python or in go or in Up and then deploy it to our platform and then use it within the spreadsheet cell, right? So, so, for example, I built a few simple functions, like, just to get your, get up stars, or something like this, right?

You can then, you know, put in your like organization and wrote the story and then you get get your kid upstairs automatically and that basically refreshes at the moment, every 24 hours and and and then basically, you know, all of these features that you can in there. R and once you are done, you export your data and you can use a bite again and then you can maybe use another like tool to visualize your data and at the end of the day since we want to become basically a part of this

modern data stick, right? We want to also you know, want to have adoption because like at the end of the day, we are building a platform that at some point. Also goes beyond pressing. Yeah. Right pricing. Obviously if you think about the spreadsheet like Excel spreadsheet, People are not using an Excel spreadsheet just to do pricing. Do a dream or yeah runs? Yeah. Yeah. It writes businesses. Right. So you can use this for CM you can use it for. I don't know, Supply Chain

management, right? You can use it for for anything that you can think of it, right? And and basically, you know, we are we are building a syntax that is very close to excel but it's still different. And also bearing functionalities that are like Like different, right? So for example, when you, when you are done on a cool shit, they have the concept of apps script, right? So, you need typescript basically to write custom function. When you go to excel, you need

to write VB a right. Obviously modern modern Excel, I think also supports typescript now, but it's still not like, you know, you can choose python or go, especially python is very important for the, for the data space because this is actually the, the one, the first programming language that every data, you know, Petitioner would go to when you go into this field, right? And, and yeah, therefore, when we do open sores, you know, we would like people to to adapt

our platform, right? Look so that, you know, basically, because if you are in a closed form Solutions, it's very challenging, right? Because people, you know, wow, you need to pay for that to use this, you know? And and, and, and F, it's a header, right? Yeah. And For example, for me most modern like also Tech stack and also businesses they actually do open source right? So they have a open source component that makes sense to it. Exactly. But so for me the custom

functions are a great example because that kind of makes the bridge in my head to like IntelliJ where you can make your own plug-in, they're like okay we don't have it the plugin for this programming language but you can make it because everything around creating a plug-in is open source, right? And you can add that into your own idea because you can create your Plugins in a way, but for them also, everything that they

have at least as far as I know. I don't know if it's a lot of things for everything, but are you planning to open source? Like all of the things that you've built so far, or is it only like the plug in components that can kind of plug and play with your core that is still kind of closed in that way? So, so obviously, we will not open source everything. I mean, we are still business at the end of day and we have to make Me in order to to basically like continue with with the,

with the business. Yeah. What we what we will do is is basically the core, right? The core is basically like like it. So no plans, right? To have the spreadsheet around that and also like the basics, custom functionalities, right? Ideally, you know, we also like someone tones our Rebel and also can starts locally, you know, like the platform itself. After see everything like that is like business-related, right?

So first a pricing logic, we build a few for example, pricing domain specific custom functions, right? So we better for example, so we are working together with press API where we're basically, you know, you can get competitive tracking data, right? And obviously this function this kind of functionality is we will not open source because you know this is more like for our business users and we Customers. Yeah, functionality, is that it basically are critical for them.

And also at some point we also have plugins that are really, like, more for our customers, you know? So at the end for them, right now, we are working with Amazon customers and we have a lot of Amazon Integrations, right? So this kind of integration, obviously, we will not, you know, just just put it out in the wild and Fine? Because at the end of the day, yeah, I mean what do you want to do with an Amazon equation, right?

As a as a developer I think what you want to do is maybe you have a hobby project you know you want to automate yourself something or also you know maybe you're a small company and you want to write your own like simple pricing there as well, right? Yeah. And then at some point you know, you realize, oh wow, it's a lot to manage, you know, the platform because this is the thing with open source. It's open source, doesn't mean

that when companies put it out there that actually, you would, you know, use it and deploy it yourself, right? This is, this is not the point. I mean, for example, we use prefect, right? Obviously, you know, it's open source. We could, you know, we could host it yourself. You know, we could host it on our own server, you know, they're not pay to prefect at all, obviously, you know, we don't do it right? Because if I would host pretty Myself.

You know, I would have liked Justin dedicated engineer to basically, you know manage that basically every day and on a day-to-day business and just it's just it's upright.

So we our company, you know, is based a lot on on open source Technologies and some of them, you know, we obviously pay because they have managed hosting services and we are happy to do that, because every service that we need to host on our side, we have to manage that and and Dang it, it's very costly because you need devops Engineers to enter that and finally good death of Engineers, right.

It's very challenging and also I don't want to you know have a full team of Engineers here that is under. Yeah that is maintaining everything and that is also you know, on the weekend you know, needs to be on a available just as easy to fix stuff that we basically, you know, have a service provider that could do it for us. And that's the thing. So I think If some people, I mean this is more like the concept for business people because some people just people just staying.

So if you put it open sauce, you know, you could just take it and and, and then that's it. And then basic redeployed yourself and not really pay but they haven't they haven't thought it through right? Because what I just said, you know, it's very costly. Yeah and I mean it's a it's a

risk, right? That is not core of your business. Like that's an add-on and it's way easier to get it off the shelf from someone that's already hosting it. And Just take it as a service, right? Its value. Both ways. If you take it on your own you also get the risk of running it on your own and your user group is just you right there. User group is way bigger, so they accommodate for that as well, with regards to security, and risk management and uptime and everything.

If you take all of that in house, then you're all of a sudden, you're trying to run their company, but not next to your own company. Like that doesn't make any sense, especially if you have multiple of those. Yeah, definitely. I mean, all the points that you set, right? So security, it's a big. Fate doesn't mention, you know, you're running our platform in the future by yourself and I mean we are running on highly protected servers right behind a V PC.

And if you need to manage yourself and if you have a data breach, right, you basically have a, you know, kind of a lawsuit against. Yeah. And then you know maybe your business is is not pricing. Your business is really just an e-commerce shop and basically, you know, you have more problems with that. Yeah, it makes a little sense. I mean, a lot of the developer

tools that I've used out there, they are open source and they call themselves opencore as well. Whereas, I think the city of Kashi court has kind of this strategy mindset in there that he says, if it's for developers, if it adds value to engineers, then it should be open source.

And if it really adds value only when it's an interpreter Enterprise, you can put a price point on it because then yeah from an Enterprise point of view you have different business models and then all of a sudden it makes sense to use this open source differently than Individual would if you talk about a developer engineer so that's also if you look at Doc or even Hoshi Corp and a lot of the other like developer tools out there they have an Enterprise price point and

otherwise it's open and it is also open source. So then all of a sudden you have this community around it that

can already adopt that. And that's also how it then penetrates probably the B2B Market, which is to meet an interesting model because we didn't use to have that, like, it was licensing or something that red hat day where it's just like offering support and and basically Contracting I guess when setting things up like this is a new business model that is more out there and more up incoming.

Yeah I mean if you just think about that new business model it's actually the business model to go because when you think about like a close Solutions right now like for example, if DACA would be close to lucian's, I don't think that, you know, everyone would use Docker now, you know, and have DACA everywhere. The same thing for a check-up. So we also use I'll form a lot, right? I don't think that I would use terraform or adapter form, if that would be like, for me, it goes solution.

Yeah. It and I would try, I'm a, I'm a developer so at the end of day as well. So I would like to try out things right and see if it's like technically like good and especially you know if it's a

community behind that, right? That would drive like kind of the development because in coastal erosions you have you have a small team you know you don't really have the feedback that Basically you would get in the in the white in the open because when you go open, you know, you just get much more

feedback. You know, you also get contribution at some point if people would use it and if people, you know developer with see the value out of it, then you could just also get things on your platform that you haven't thought about, or maybe also people will use it in a way that you have a thought about that, right? And so, you get a lot of ideas, how people would use technology. And at the end of the day, we also, Creating a tool stack,

right? And we want to see how busy at the end of Day Developers would use our tool. Yeah. There's so much value that comes with opening it right? And also I mean if I see a price point on a product, I want to use, that's such a hurdle. I'm like I'll just try an alternative. Like I'm not going to pay for that stuff.

Even if it's like a trial thing I'm just gonna I'll move on to something else where I don't pay it and I can just kind of fiddle with it because that at the end of the day like from an engineering point of view, you try a lot of options and you also Also, try the options that are most accessible to you. And from there you make your, your pros and cons, right?

And you pick based on that. But if you're not even in the running, if you're not even an option because there's a price wall there yet, then you're never going to get adopted as widely as Docker is or terraform is or like any other tool that's in the kind of developer tool belt nowadays. Yeah, it's an interesting thing. I was wondering because you have

the, the CTO role in. You also come from a developer background. I was wondering like, how much is your time still spent on Hands-On coding and how much is it more? So, like the CTO roles and responsibilities that you have now? Yeah, so I mean, this is probably that a lot of people ask. So have you have you stopped coding wets, you become management and she quite strongly happy spot because we were still a small team. So we are not like What I do + Engineers or something like

that. Yeah, so we're like 12 engineers and spread between two teams, and I'm so pretty much involved into coding. And so I still also, you know, if the team has some capacity issues, you know, I jump in and add two features but I also, you know, just have to team to debug certain stuff. So kind of like, you know, the guy for like the Swiss knife and To do everything around this and

that and yeah. But obviously what I need to do to balance is also like the the time that I spent a coding, I spend less time basically on product development, you know, and also so aligning. So I think when I started with price to obviously it was a lot of coding in the beginning because, you know, it was quite quite great new, you know, we didn't have When so, I still, you know, get my hands dirty, a lot of coding. Then basically, when the team has hired, I reduce it to quite a lot.

So it was more than, you know, like, like getting the vision right getting the requirements, you know, creating the culture hiring obviously, right product development. But, you know, I'll everything around that sense of like, you know, you do present, a user research, talking to customers and then, and then basically, Lee also getting the team on board like because obviously, when you start, you still don't know what you are getting interested.

Yeah, you need to basically have a good onboarding for everyone and then and then basically, when this team said it's in, I, you know, started to cook more as well, right? So based on the time that I have and now I think it's a balance so I would say that I code ten to fifty percent of my time. Yeah. And Coding is just not always just, you know, create new features. But I said, like, also, bug fixing into a nothing arrest.

And everything else is basically like management roles and also communication between teams managing teams. Yeah, try to make everyone happy. That's awesome. That you highlight that. It's like, it's a growing position, right? And you start off, obviously in MVP phase, you have a lot of Hands-On work to do because you want to deliver that as fast as

possible. But as you Onboard more people, you roles, and responsibilities change and all of a sudden other like the things that you're accountable for our more highlighted, right? And are more important than you being Hands-On on a coat on a keyboard basically contributing in that way, which is really cool because I always thought like, I mean that's very much so because it's in a start-up guy I think in big organizations to bit more tried and true.

But I like the Highlight that it's just a wild west sometimes in a start-up and it really depends on the context and the situation. Where your value is. And I think a great CTO really just has to foresee the value that is there. And also then be like, okay what are the trade-offs here? What can I delegate and what do I need to do in this position now? I think that's really cool. Yeah, definitely. Yeah, I was as a final thought I still had because I always think

this is interesting. The first hire was that like pre

MVP or was that post MVP and how they like hiring that first person or on bordering the first person go. And so, If hiring the first person was actually pretty and BP. So I think as a we started in October 20, 20. Yeah. And our first Hayek was already in think end of November starting of December, if I remember correctly. Pretty slick was very quickly at the first. I actually was coming through my personal Network, so he used to

work for me before. So, therefore, it was not really like difficult to Do to Bob. Because to be honest hiring, if you, if you are like a first-time CTO, if you don't have a network before, I think hiring is very challenging, like, especially if you don't have a network, like, it's like, finding good Engineers or finding like, you know, someone like at that point. Because obviously, when you work for a start-up, it's different when you work for a corporation.

Yeah, because in a start-up, there's a lot of Chaos in the beginning. It's very unclear where to go and it's very Radical right? So, whereas when you would be in a corporation, it's more settled. You know, you have obviously I mean hopefully you have Revenue. Hopefully, you're not making loss and son. And you have a stable product that uses a use already on a

daily basis, right? When you, when you first higher, you know, a higher that you need to basically hire people that are very suited to startups because because many people are not really used to this. Go change and because in a start-up, you know, from one region to another, you know, priorities by changed completely. Yeah. And I've seen some of the passengers. I would say if I look at this like backwards, they were not really used to that like, you

know, this radical change. So they were really frustrated with basically, I know, you know, you need to give me things that, you know, stays for three weeks or longer because this is what I'm used to write. Constable. Yeah. Yeah. But this is, this is, this is like the corporate mode, right? Exactly covered moment. You know, you have people who, you know, they they need security, right?

So they want to know what is basically the goal of this quarter and you know it has to be defined very very like you know you know in a waterfall level completely and then you're going to work on that right? In a start-up obviously you also have a goal what you want to achieve in the quarter. But reaching this this this this goal, you know, can can go Go from my day to day, from week, to week, and until you have something more concrete, right? But I think that is the fun out

of this, right? That you are actually looking, you know, for the solution and not really like, you know, you have a solution already and you just come in and then just pull it because then you know, then then then it sort of start up exactly. Like it's a it's a different type of fulfillment and I think the highs in a start-up, I'm going to be way higher than maybe a traditional corporate job would be but there's also yeah there's loads Those that

are going to be lower. Probably in there's high-risk and high-reward in there as well, but that probably makes it just a fun journey to be on the yeah. Definitely. How did you get your network?

Because you mentioned that as kind of the key factors in hiring that first person, how did you get your network up to a point where that was just like a comfortable higher for you as a first timer? And I mean I mean I obviously I before phone and press through this was not my first job so I was experience. I already had some experience. I used to work for acts of spring. It's a media company I had AI division there.

I was a Diallo before and I also started to know career my career at to Votto like a u.s. suffer company. So they are there are no part of VMware but basically, you know, over that time, I already started to, you know, create a network. Yeah. So I always, you know, love writing. So I always chat Intent. That was actually not intended to grow the network because I guess some people always ask me what is your strategy to increase the network?

I also want to, you know, increase my LinkedIn account and all of the stuff. And I said, yeah, you know, I didn't really have a stretchy before. Then I was like, you know, I actually just took it kind of like to, to remember what I mean I read. Right? And I and people just just like it and so on. And also, you know, I love

writing and brought posting. So at the time I visit, you know, I was writing up my knowledge that I learned so they can you know, remember that and obviously you know okay cool. If I already write it you know I can share it there here and there and yeah people liked it. Yeah. So and then I also, you know, saw the value and Open sauce because you know for now, for example, when I started it Lydia low. It was actually my first head of position. So my first leadership position

and I was pretty young. So I see that was I only had two and a half years experience, and it was ahead of data product roll. Yeah. And actually, the thing is, I applied for that role? Yeah. So the it's a funny story because the role requirements was like they were looking for someone eight plus years experience. Yeah. You know, have have met, it's, you know, lots of teams and do that. I didn't do any of this. I was I was, I was like, I was

really bored, you know. It was like, you know, I was like, I was showing them what I did, you know, at pivotal that kind of, you know, would also show that I've managed teams also, obviously, doing University time I was also leading organizations and everything has already, right? So it's not like to know I never had leadership positions within but not you know, in In a professional context. Yeah and and then you know I applied for it and I got a I got

an invite. So I talked to the CTO and he liked me and then I actually was able to pitch my ideas to the co-write to the managing director at the time. And yeah I got it. I got the position even though even though you know the the toughest person was completely off from from from from my level and everything else. And and basically, you know, at the time of that joint.

Yeah, I know I also was like one of the strategy was to do open source, yellow hasn't done open source like a proper strategy, right? Because because obviously when I started at the yellow I didn't had a big Network before. So the network was quite fine. You know it I think I had like five thousand or something follows already, you know because of my blog posts and so on but not the amount that I had to find also not the There. Come the people.

You know, because I didn't manage people so far and then I started. Okay, I need to create content. That makes sense for people to join ideologue and also projects where I can actually have a sustainable way of basically maintaining, you know, the team in some way, right? Because at the giallo the company was 18 years old. It's bitter, I mean it's 18 years, sounds pretty young but they're, you know, the whole Tech stack and things like that.

Was pretty old-fashioned. So wasn't really modern. I think, I mean, at that time, they were still running a lot of stuff on Pearl and it just started to migrate to, to Spring boot, right? And, and for me, open sauce, you know, was one of the core pillar to the strategy to maintain, also, to open up the company itself, right? And so, a new data was important, but I also wanted to help the company, you know, to become modern right in a way. So I create, The idea low-tech block, right?

I created the trailer Channel and everything else. So this is funny because you always think HR or someone else would do it but actually it was coming from me and I crazy you know ecosystem around that, right? So open source, which Drive the blog post blog post will drive Open sauce, blog post will drive Twitter going to conferences with drive all of this together, right? And and and over the time, you know, became also vary Popular.

So, I mean, one of the projects like image, stood up that as described at the beginning, got four thousand stars like, you know, like, it's so, it's actually, it's actually here like something that was part, you know, when I started doing that at the time. Yeah, yeah. I love hearing that building your network.

Can go on various ways, right? But if that's your main goal, then sure that's going to be one of your goals, but the best thing is, if you love what you're doing in through that, you can build your network, right? And that can be through putting Ting out content that can be through gaining a position by being bold in that way as well. And picking up things to add value to a company. While also extending your personal Network. I really love those examples that I had a lot of fun covering

your journey through price. Loop your past as well and you're taken. Open sourcing. Is there anything that's still missing? That you would still like to share with our audience before you round off? No. So thanks Patrick for inviting me. So, I had a nice chat with you and yeah, I think so. We Quite a bit of things and hopefully it might be also interesting for potential Founders, you know, or people

who also tech-savvy, right? And who wants to found or like you know, be in the same Journey as me in the future because at the end of the day I still think that there are too few like Tech Founders, right? Who basically do this in Europe especially so the founding seen in Europe is very driven by business and Business Schools. And I actually would like to see more like Tech Founders like really Tech, Founders, right? Who would also shape kind of the future of technology here in

Europe and create? Yeah, unicorns. Yeah, I love that. No. I apparently, they're all in Tel. Aviv IRAs. Me find out if I don't know, but we're gonna round it off here then. Thank you so much for listening. I'm gonna put all that socials in the description below. Check him out. If you're one of those Tech Founders and you want to come on and hit. Us up, will make an episode happen and in any case, thank you for listening. I will see you in the next one.