278. Meet The Exited Founder Building Infrastructure for Adaptive Software w/ Iris ten Teije | Sky Valley

0:00Amardeep Parmar:Iris inside out. So you came on here before in your previous company, Marva, and you've since exited that and you're building something new. Can you tell us about the exit you had and how that all worked out and came into place? Sure. So just as a bit of context, the last company I built was a SaaS company. We built a customer support platform, specifically working with companies that have large online communities. So we had a lot of clients in the crypto space, some in the gaming space, some in the AI space. So we had been running that company for about two and a half, three years. We were at a stage where it was growing, it was profitable.

0:34However, we also felt we hit a bit of a ceiling in terms of, okay, this company still has potential and can still grow further. However, we're not quite sure that this has VC skill potential. So we would rather move on and do something else. And yeah, I managed to sell that at the beginning of this year to a firm called Nusa Labs. and what they do is they buy and operate software companies. So they're still running the company and they are aiming to grow it and increase the revenues. And, you know, obviously myself and my co-founders helped them during the transition, but now they are fully running it and yeah, I'm working on a new business.

1:14Amardeep Parmar:And with the exit as well, so obviously we've been working for several years to get to that point. How did it feel when you finally got that over the line and was like, okay, now this is done and it's a new chapter to begin? I think with what any finder would probably say it's it's mixed feelings. On the one end you know I was really happy with the fact that the company is still you know living on and because I think it's always a bit sad if you build something and then you have to shut it down or something like I'm really I'm really glad that it's still you know living on. However at the same time it's a bit of mixed feelings because going in obviously we had huge ambitions that we weren't able to like totally fulfill in that business.

1:57Amardeep Parmar:And what did you do after you took a bit of a break right? Uh yeah, I traveled for a couple of months, sort of was thinking about what to do next. Pretty quickly realized like I really want to start another company in general because I like building things but also looking at the market at the moment there's just so many opportunities, so many things are changing so I think this is a really good time to to build um so yeah I knew I wanted to build something again and then it was just a matter of finding the right people to to build with and finding the right problems and ideas to work on and then like when you're getting from like obviously working really hard to get started with the ground to then having that rest time like were you able to like switch off and relax was it straight away thinking about I said that next idea?

2:42No, I was able to switch off for a little bit, but yeah, it's probably not something that I could do for a long time because you do get bored and I just want to jump back in.

2:52Amardeep Parmar:And you said like you found a co-founder, right, as well. How did you find a co-founder? Yeah, so when I decided that, okay, I want to build something new, I knew pretty quickly that my preference is to work with someone, one or two co-founders, just because, I mean, it's a personal preference but I think you know building a startup it's so difficult both in terms of it's a lot of work so if you have multiple people you can just do more but also mentally um you know you're going to have good days bad days and I think being in there together with someone is just hugely beneficial not just for the productivity and what you're building but also just for your own sort of mental health at least for me um so yeah I started looking for people to work with tapped into like several networks my own of course my personal network but then ultimately ended up finding Noam who I'm currently working with through the YC co-founder matching platform and yeah generally just as a tip I guess for people listening or looking for co-founders definitely would recommend it because aside from Noam I also met a lot of other you know really amazing great great people through the platform and the reason that I ended up working with Noam and not necessarily with the other people I met, they were also like, yeah, some really great profiles.

4:08But yeah, it's always about, okay, what are you interested in? Aside from compatibility, what do you want to work on? Yeah, having similar interests.

4:19Amardeep Parmar:And you said as well, you kept a good relationship with your previous co-founders as well, right? And it's just different life stages. So it wasn't like there was a big breakup or anything like that. It's just working out the next chapter and how that goes as well. but when you look at say like when you were looking for your co-founders your previous company to having built a company for several years then looking again was anything you looked for different or you were looking for in your new co-founder that maybe you didn't place much important than before so when i started my previous company that was the first time i i had already worked in early stage startups and been a first employee so kind of knew the startup journey but it was the first time being being a founder and at that time i actually joined antler so in some sense they do prepare you quite a lot in terms of there's a lot of content and workshops about these are all the things that you should be mindful of when looking for a co-founder ask your co-founder these 50 questions so i think i had a bit more context than someone perhaps just going in in blank so that really helped at the time um so even though i didn't find my co-founders through antler i would say that when i was looking the last time already i was quite you know mindful conscious of in terms of like these are the things that that are important to me some things are still are still are still very similar so obvious the obvious ones like complementary skills what do you want to work on also like some practical elements like what what's your personal runway like do you need to raise money right away or can you work on this thing for six to twelve months even if we don't raise money like right away um i would say one of the things that really appealed to me about Noam and that I wasn't thinking maybe so much about the last time that I was building is at that time I was thinking more about complementary skills in terms of hard skills okay I've done marketing I've done BD and I'm not a developer so I want to work with a developer this is still important but I was thinking now also a bit more about complementary soft skills or like personalities in a sense of I would say like for myself I am very pragmatic and I'm always thinking okay how do I get from zero to one what what can we do today to get to our to hit our next sort of milestone um which is important because you need to keep things moving however what I really like about Noam is that he is more of a um he has more of a visionary style and the first time we met he immediately led with this is what I want to this is what I want the future to look like in five to ten years and the first time I met him I didn't understand all of his ideas but that actually was appealing to me because I think it's a good you know it's a good balance where I can think more about the sort of practical day-to-day whereas now I'll make sure that we don't lose sight of the big vision and so the previous startups obviously you said about a company skill sets but you probably learned quite a lot through that previous startup too and what is it that like today you think like actually is really the strength you can take away and the learns you can take away from the first startup that you may be more confident in certain skills than you had before uh yeah sure um in terms of not necessarily skills but just learnings in general and things that i would do a little bit differently this time.

7:36I think one of the things and I mentioned it to you just before recording is that you know as a second time founder you're very harsh on yourself and your own ideas and you're not you wouldn't like dilute yourself with okay there's a bit of traction you know let's just keep trying for a bit longer and you're kind of just hoping this thing will work out And of course, if you try longer, you can probably get some traction, but it's probably not going to be good enough. So in the end, you've just wasted that time, those weeks or months or maybe years, depending on how long you keep going. So also now that, yeah, in the current company, so we've been working on this for a couple of months, it's still pretty fresh, but we've already been experimenting with four products.

8:22Of course, all of those products are somewhat related. It's not totally unrelated random products. But we're just testing things and very quickly making a decision, okay, this is nice, people kind of like it, but it's probably not good enough.

8:36Amardeep Parmar:And what kind of area are you testing in right now? So what are some of these experiments you've tried? Yeah, so in terms of what we're building now, so our company is called Sky Valley, and our product is called Ziffer. And what we're building is infrastructure for dynamic software. and what we mean by that is software that can adapt as the user is interacting with it. So that's quite different from how software behaves today because most software today, a developer codes it, you deploy it, then if you want to make changes, you kind of go through that same cycle again. Of course, now there's AI, so you can code a lot more quickly, but you still always go through that same cycle, whereas what we're thinking is that in the future, that will no longer be necessary because some software can adapt at runtime as users interacting with it.

9:24And we've been seeing a couple of small examples. It's a pretty new area. We've been seeing a couple of small examples from some of the big AI labs and Google. So, for example, Antropic, a month or so ago, they released a demo called Imagine with Cloud. And how it worked is that you would write a prompt like you're used to when you're interacting with AI that would spin up a page. but then what would happen is if you click a button on the page a new element would appear and new elements would keep appearing just based on your interactions with the page without you explicitly prompting for that change um also google last week actually they came out of a new research paper about beyond like text ai chatbots and ai interactions beyond text because today most interactions when you talk when we talk to an ai chatbot the result is text and And it doesn't always make sense.

10:18If you want to explain like a complicated concept, maybe it's better to create a web page and show diagrams. And so we're thinking these types of experiences are going to be much more common in the future where software just gets created on the fly dynamically as and when the user needs it. So that's the longer term vision that we're working towards. And then in terms of the short term products that we've been or the smaller experiments and products that we've been working on today. So they're all sort of related to one core concept, which is around storing and understanding intent. Because our belief is that if we want to create this dynamic software future, the software cannot just change in any direction.

11:05There still has to be some guardrails and principles and it has to follow the developers or creators' intent. So we have to understand what was the developer trying to achieve when they were trying to build this software and how do we capture that intent. So that's the first building block that we're working towards. And you can think about, for example, storing prompts and conversations that you have when you are building something with cloud code or codex. So that's actually something that we're working on today. We've already sort of built that. So, okay, how can we store all of those interactions?

11:44make some sense out of it so that we understand the full context of what you were building when you were building it and not just have access to the static code which is more like the outcome.

11:57Amardeep Parmar:You mentioned like Google and Anthropik there and people like that as well and as you're kind of building up this future is this something you're building a collaborative with these other organizations is it how do you kind of would you fit into a tech stack for like a future developer? Yeah, so where we see the value and where we're building is we want to become the new system of records. So the place where software creators would store their work and the work can be the code, but also as I mentioned, the intent. So these conversations, these prompts, like what they were trying to build when they were building this piece of software.

12:34We don't necessarily need to build everything ourselves. So there's already a few companies out there, for example, there's a company called Thesis with a Y who's doing like gen UI APIs. And for example, we can definitely collaborate with companies like them. And, you know, the other big labs might also be releasing some APIs that will make it easier to integrate those types of experiences within software. And again, there's no point or no need for us to compete with that. but we still think, okay, so there's still going to be a need to have this system of record and this place where you store your work, where you collaborate on your work, that will have to be integrated.

13:14So you would need both the code and the intent, but we also have to be aware of what's happening as the application is running. So we have to make sure that we can combine both those elements.

13:30Amardeep Parmar:Hello, hello. I hope you've enjoyed the show so far. I'm Amarit Parma, co-founder of BayHQ and the host of this show. For those of you who don't know, BayHQ is the community for high growth Asian heritage founders and investors in the UK. Over 7 ,500 people have attended our events, 200 people have been for our impact programs, and obviously there's been over 250 episodes of this podcast. If you want to join us and take part in the program and come to the events, go to fahihq.com forward slash join. We also just released our first ever book called Startups for Outsiders, which you can get on Amazon now and the link is in the bio.

14:11Amardeep Parmar:Hope to see you soon at our event. Hope you enjoy the rest of the episode. I think like people didn't obviously realize this sums up. So I used to be a developer myself, right, back in the day. And it was the early days of Oracle chatbots, for example. And there's a whole thing about finding that intent. So people often try to do it for that consumer side and the customer side of what's somebody talking to a chatbot, what do they want, right? But with what you're doing on the developer side, obviously, it's going to be a lot more complicated. But then also the potential value there is huge, right?

14:39Amardeep Parmar:And obviously, it's about a five to 10-year vision. Like, what can you imagine seeing in that 10 years time of if everything goes to plan and you are able to achieve this? So there's a few different use cases. And I think that's also something that we cannot imagine today. But what's the point of this is that software is going to be a lot more personalized. And you can just spin up things on the fly as and when you need it. And this can apply to B2B software, can apply to B2C software. So let's say in B2B, you've got a dashboard of analytics. And quite often, there's so much data to tap into. do.

15:15So instead of why would you want to pre-code five charts, if you can just show the data that is relevant in the way that is relevant at that time for that specific persona, that will be possible. Also, for example, you've got CRMs, typically they're very complex. There's so many fields and you don't use like 90 % of it. So why do you need to look at all of these things? And it's kind of confusing. The software ideally should just adapt to you and removing all of the unnecessary complexity or when you're running into an error, does it always need to go back to the developer for the bug to be resolved or can it so far heal itself while it's running?

16:04So those are some of the areas that we're thinking about. And like a lot of these things, they're not necessarily going to be visible to users. at least the way that I'm thinking about it is more like for example if you think about Google like if you're searching for something let's say you're googling for like restaurant recommendations and you're in London you're going to see results that are in London and that just makes sense and you don't necessarily think about that it's like magical it just makes sense it's relevant it's contextual and software is going to be like that too but today it's too expensive for most companies to make the experience so personalized for every single user, for every single use case, edge case.

16:46It just doesn't make sense if you have to code that into like a static system.

16:51Amardeep Parmar:And then as you're building this out right as well, so you said you're going for the experimental approach right now. Then let's say you're doing the different experiments, is then to build on top of that? Or how would you kind of see the roadmap coming to your eventual goal? Is it doing different experiments, then kind of doubling down on an experiment going well? or is it these kind of experiments aren't necessarily what's going to be the final product? So we know, for example, the storing intent is going to be crucial regardless. So we're doing that and then we're just seeing, okay, what can we do with that product today?

17:22Which is not, that product doesn't fulfill our entire vision. It's just like a small part. But can we already, is this product already useful for a small group of people? For example, in our case, we've also been building a community of vibe coders can they use that product today for some of the vibe coding that they're doing even though there's nothing to do with dynamic software necessarily and obviously because you had the community

17:45Amardeep Parmar:platform before right so you're you're going to be pretty good at bringing a community and like being able to yeah i'm very bullish on community um even though i mentioned that previous company wasn't maybe not maybe not necessarily vc skill i am still very excited about communities building communities the first thing I did when I started this new company even before sort of starting to build product is reaching out to people in the space and having having a lot of conversations and once I felt okay now I now I've had enough one-to-one conversations that I can start a community we started it was just of a simple whatsapp group so also if anyone you know watching is interested and into vibe coding to reach out to me it's been a really it's been a really great group actually in terms of people supporting each other asking each other for advice on their projects and especially in the community is important anyway but especially in this early stage I really like it because when we have new ideas I want to get some feedback I'm just very quick you know I just message the whatsapp and immediately I get that feedback okay maybe I should explore this more maybe not and then you were saying earlier as well how you think about funding differently this time compared to last time too how you think about that funding element or doing this company?

18:57Yeah, so for us, we definitely need to raise because what we want to build is going to be quite resource intensive. But we only want to raise once we feel, okay, we've got a bit of traction because otherwise we might get into this trap where we're raising based on a specific pitch and pitching a specific product. and then if we want to you know experiment more we can still do that of course after we raise but it's different if you already sort of committed to your investors or told your investors that this is what you're going to do even if they're totally fine with with the founders changing course like from a mentally it feels it feels different if you already taken money versus if you're

19:45Amardeep Parmar:still bootstrapped and in terms of geographics as well so both you and your founder are both place in London right how you think about how you expand this globally or is it gonna be very UK so even like now for our vibe coding community it's a little bit more London heavy just because we're here but already it's global and yeah not a surprise it's mostly like US and then London and yeah in terms of this company for sure longer term probably the US is going to be the biggest markets and obviously I think as we're recording this right now the budget is coming out and And there's been a lot of talk at the moment about like UK as a place to build and obviously built your PS4 company here as well.

20:25Amardeep Parmar:What is it about the UK or London that keeps you building here? I think a few things. There is a lot of good talent here in terms of, yeah, actually in every area of every part of the stack. And then also, you know, if you've been in a certain place for a couple of years, you've built a strong network. and that is actually one of the things that that's kept me here both from a sort of personal point of view but also obviously from a business point of view like now if I want I want to if I want some feedback help if I want to raise I can just reach out to all of the investors that I've spoken with before so it's just so much faster and yeah as you look to grow this now so have you at the moment is it the two of you or you started to hire or how you thinking about growing that team too because obviously with AI experts I'm guessing you're kind of making use of agentic AI and things like that too so how do you think about how you're going to hire and grow?

21:20I think like as any startup everyone now has the ambition to build a billion dollar company with a team that's as small as possible so yeah no difference for us so at the moment it's just myself and my co-founder given the AI given the possibilities of AI and the fact that we're also working with a lot of vibe coders I personally also got really into it so So these days I'm spending more than 50 % of my time actually building, even though before this I've never been a developer or never, yeah, the most I've done is like build web flow sites and like automations, but I never really built, you know, web apps or mobile apps.

22:01And now I am. So I'm spending a lot of time on that. But we do, once we start raising, we do, of course, need to hire people, more engineers, community marketing, but also probably our first hire is going to be a designer because I think that design is really important. It can be a strong differentiator, especially at this time where anyone can build anything super quickly. It's really important to us.

22:29Amardeep Parmar:And I guess on that side as well, so obviously having built communities before and how are you thinking about building that up? You said you've got the community of Vibe Creators that gives you direct access to people how come how you're going to protect this as so many people are exploring this space right now like what things are you doing to kind of keep that traction and that momentum yeah i think with community you have to be very very personal um you have to be helpful like you cannot start a community with thinking about i'm just going to push my product or promote my product and it sounds obvious but i'm see i'm still seeing there's a lot like people starting Discord servers and just writing product updates and people just aren't that interested in it.

23:13The primary goal of the community should be people getting access to resources, knowledge, helping each other. And then, of course, you can talk about your product from time to time. And also, I guess one of the metrics that I always have for really good communities is that people build relationships between each other. Okay, they find co-founders or they find clients, that's what makes communities really sticky and good.

23:42Amardeep Parmar:Yeah, it's one of the things that we're trying to work on our transition as well, because I'm kind of the face of the community at the moment, right? But the whole point of the community is it shouldn't be a cult. It's not about talking to me, it's about talking to each other, as you said, and now that's starting to happen more. So I'm trying to slowly, like, the more I kind of disappear a bit, it means that hopefully then people start helping each other and then hopefully I can just sit on the beach and not have to do anything people that that's the aim right um well I can just do the bits I like to do rather than having to like do all the communication side and let's say like what we're going to do now so you've been on the podcast before but if we get you on again and say 200 episodes time which is about two years time where hopefully we're going to jazz up this area of this we never run we're going to have like a tv up here and we'll play this back in two years time when you come on again about like where do hope that in two years time you can say like this is our biggest achievement so far?

24:32I would want to be in a stage where we've got serious skill ups and also enterprises use our infrastructure to build these dynamic experiences.

24:45Amardeep Parmar:So we'll play that back when you come back on again and then see hopefully you've been able to achieve that and you can sit here and be like, yep, we did that. So before we go to a quick five questions as well now, what was it just about your growth? And it sounds like this kind of new chapter has been really exciting for you as well. What are you enjoying most about this new startup? I would say the fact that I got so into via coding and yeah, that's been really life-changing because I think every commercial founder has had that frustration in the past. you have so many ideas but you can't really execute on all your ideas because you've only got if you're lucky you've got a technical co-founder but the technical co-founder can't work fast enough to make all your ideas happen and now you can do a lot of that yourself and you also i think as before we started recording said about how one of the things with your co-founders off because they've also previously started a company they understand the hard parts and the the struggles in the journey too and like what keeps you going what keeps you when you know how difficult things can be when things go wrong what kind of motivates you to keep going and keep pushing new frontiers it's a few things i would say i would say for this specific company to me the fact that this future of dynamic adaptive software it's inevitable so i know it's going to happen and i feel i have the chance now to be like a core part of that if i give up it's going to happen anyway and i will always wonder, oh, why didn't I do it?

26:17Amardeep Parmar:Awesome. So I'm going to go to quick fire questions now. So you've been on it before, so I think you know the drill. So first one is who are free Asians in Britain you think could do an amazing work right now and you'd love to shout them out. Yeah. So I'm going to mention some founders who just recently started new projects. So hopefully some new companies to check out. So one of my friends, Alex, she recently launched a company called Hula, which aims to bring new parents, new moms together. So it's AI and Second Aid. Alex Ashwood, yes, but also very focused on real life community. And that's actually also an area that I'm quite bullish about.

26:58Before I started this company, I was thinking about, do I want to do something? Because I'm really passionate about community. Do I want to do something in the physical community space? I ended up not doing that, but I still think that's a really interesting space. Secondly, YU and Z, who recently started a rec tech company called Proximity. So they're a very strong team from a fintech and technology development background. And they recently got into YC as well in their latest batch. So check them out. And then lastly, I'd mentioned my friend, Parit, who is building something in stealth in the cybersecurity and visibility in the AI space.

27:44Amardeep Parmar:Awesome. So check them all out. And if people want to find out more about you and what you're building now. Yes. So you can check out our website. That is getdiffer.com. And people want to follow you? Where's the best platform for you? Yeah, LinkedIn or X. So that's my name, Iris Tantaye. and I'm sure we'll put the links in the description. And if anybody is listening now, could potentially help you, what could they reach out and help you about? I would say two areas. One, if you're really into Vibe Coding and want to join our community, give me a shout. Or two, if you are interested or have any specific opinions or insights around the dynamics of our future, also love to chat.