Episode No:100

Boosting Collaboration and Efficiency: Improving Requirement Management

Mikus Krams

Co-founder, COO at Trace Space

Listen & Subscribe


Ep#100: Boosting Collaboration and Efficiency:
Improving Requirement Management with Mikus Krams (Co-Founder & COO, Trace.Space)
Ep#100: Boosting Collaboration and Efficiency: Improving Requirement Management with Mikus Krams (Co-Founder & COO, Trace.Space)
  • Ep#100: Boosting Collaboration and Efficiency: Improving Requirement Management with Mikus Krams (Co-Founder & COO, Trace.Space)

Episode Summary

In this episode of The Hyperengage Podcast, Taylor Kenerson and Adil Saleh engage in a comprehensive discussion with Mikus Krams, the co-founder and COO of Trace.Space. The conversation centers around Trace. Space’s innovative requirement management tool, designed to modernize and streamline processes in sectors such as automotive, defense, and manufacturing. Krams also shares insights from his prior experience at Chili Piper and his roles in mentoring and angel investing. A significant focus of the episode is the customer-centric approach adopted by Trace. Space, emphasizing the importance of understanding and addressing customer needs, whether they are common requests or unique situations. The team also delves into the challenges of managing customer expectations around AI capabilities, underscoring the need for setting realistic goals and the limitations of AI without proper data training. Furthermore, the podcast highlights the company’s commitment to swift responses for bug fixes and the crucial role of continuous product refinement based on customer feedback, illustrating Trace. Space’s dedication to both innovation and customer satisfaction.
Key Takeaways Time
Identifying the generational gap between legacy requirement management systems and the expectations of modern software. 1:23
Validating the problem by getting feedback from 10+ senior systems engineers 2:02
Maintaining close communication with early customers to truly understand their problems 3:43
Resetting customer expectations around AI capabilities 7:51
Differences in developing hardware vs. software tools 14:14
Ensuring quick time-to-value for customers 19:35

Empower Your GTM Strategy

Monthly expert advice and top GTM insights in your inbox.


[00:00:04] Taylor Kenerson: Hello, everyone. Thank you so much for joining us today on another episode of the Hyberengage podcast. So thankful to have you here in 2024, and here's an amazing year. I'm Taylor with my co-host Adil and a beautiful guest, Minkus, who is the cofounder and COO of TraceSpace. It's a requirement management tool, And he also has a wealth of experience previously, which we'll dive into a bit. [00:00:30] Taylor Kenerson: He was the director of strategy and finance at Chili Piper and also previously had dabbled in some mentoring and angel investing, which we'll definitely get into. So Thank you so much, Minkus, for joining us. We really appreciate you. [00:00:46] Taylor Kenerson: Thanks for having me. As some of us as some of us know and maybe some of our audience. We know that requirement management has been a big issue. So can you kinda, like, walk us through your journey of how you found this problem and kinda how you went from 0 to 1. [00:01:03] Mikus Krams: Yeah. Sure. So actually with, so my other 2 co founders, been quite deep into this. And, they either worked with competitors or they've been implementing these competitors. And so one and then with one of them, I was working together with Localize, which is kind of a more modern b2b SaaS, solution. [00:01:23] Mikus Krams: And he really saw that there's this kind of generational shift from what's out there in the market and then kinda what episode. Software, it looks like today, what kind of experience people are expecting. And as we looked at the market, we just saw that there's really really, it's sort of really behind and then kind of stuck in this gen 1.5, you know, digital transformation going from paper to software. But it's sort of stopped there. And if you look at tools like, you know, Notion, Figma, and so on, it's just much more pleasant, fast, collaborative experience. [00:01:56] Mikus Krams: And, and so I was the last one to be convinced that this is a good market and kind of I asked my cofounder that, look. We need to find At least 10 senior systems engineers that, unprompted, tell us that this is something that they're struggling with and they need help with. And once we got to that point, then we went out to raise our pre seed, got the money, and started building, and here we are. [00:02:24] Adil Saleh: Very interesting. When I first looked at, phrase space, you know, previously, you know, background with talking to a lot of, You know, feedback management platform, more built for, you know, customer facing teams, product managers, and all. And, we've never like this Seem like pretty unique for us to, you know, someone going out and building it for for for automotive, for industries like defense and manufacturing, You know, where, you know, processes are so legacy, and it's so hard to get the you know, a lot of flood of the feedback gets lost in translation, or you have to A lot of time and and and and resources, and and and there's a huge impact, like, element of, you know, optimizing the cost Because the customer lifetime value is already big, but it's so hard for for for you to get the the best time to value in the in the in the beginning post onboarding. So how did you achieve it in the first click? Like, of course, building, validating it with with engineering team. [00:03:36] Taylor Kenerson: let's say, [00:03:37] Mikus Krams: first customer? Mhmm. Sure. So actually, the we started very early by validating, like, Figma designs. So even before we had a product, we were, you know, we had about 20 people that we were really close to and said, is this is this kind of what you're looking for? [00:03:50] Mikus Krams: Is this helpful? Is this not helpful? And even before we had a proper product, we had some iterations. And I think that's a big component of, you know, we just assume that we know nothing and, you know, and try to build the best possible for these systems engineers. And then as we get started with the customers, in a way, because we're so small, what we can afford is really essentially, we have a common Slack channel or WhatsApp channel or, you know, Microsoft Teams, whatever they're using. [00:04:20] Mikus Krams: And we put their whole team in the channel and our whole team in the channel. And so we can be really, really close to them. And because for our stage, it's, you know, every interaction is really valuable, And we wanna make sure that's gonna if there's if there's problems, we understand why they're running to them and so on. So that's That's how we make sure that we are aware of what's happening. We also have, you know, things like maybe some logging and so we can see, okay, you know, who's, How much how many activities people are doing and so on, but it's much more important that, we get that kind of qualitative feedback, not just a quantitative, you know, you you had, you know, you created x many items, and I think that's I've seen also in past companies really this kind of, best in class customer support That you really answered very, very quickly, and you actually resolved the problems, and you escalated very high up. [00:05:12] Mikus Krams: It's really worth it, especially in the early days, and especially if you're serving relatively small number of really, large and important customers. [00:05:21] Taylor Kenerson: That's a huge That's a huge component of, you know, a lot of companies is, especially in the early stages, a lot feel like they have to, you know, just double down on the product. I think you often miss that. Like, the customers, they're giving them this advice and this feedback. And how do you bring it into that, like, flywheel where you're able to constantly push and get just like you said, actually solve their issues, not just, you know, elevate it to support, and then where does it go? So can you briefly touch on just how you're able to prioritize your time and your resources, but also manage, like, at a very you know, with, like, love service, these customers and ensuring that they are reaching the success that you promised and, you know, they're expecting. [00:06:05] Mikus Krams: Episode. I'd say it's probably the hardest part. It's really to to to figure out, you know, what to focus on because People don't ask for the same things exactly. And, you know, if if they ask for the same things, we probably have them built already. And it's it's really trying to figure out the things that maybe one or 2 ask. [00:06:24] Mikus Krams: I think the prioritization kinda comes down to, you know, is do we foresee that more people will want this? Or is this Something that is a very, very cornered case. And a lot of times, I think, especially with AI, there's been a bit of a the challenge has slightly been that People have a lot of expectations that and they've been oversold by, you know, by Twitter and x, you know, that that there's so many things that possible. But then in reality, especially in engineering, The it's you need to be very, very exact. And all these hallucinations, all these, like, kinda probabilistic answers, they don't really work very well.  And so that's where sort of people say, can you build us this? And we say, we can try, but it's likely not going to be what you expect. But that's I think that's that is probably the biggest challenge. And really, we just try to be honest and say, look, this is how we're building. It is how we're approaching. [00:07:16] Mikus Krams: Is that really do you think that's gonna be helpful? And if not, then maybe let's not go there. On the other side, things like white gold treatment for Boggs is just that, you know, we Somebody says, hey. Something's wrong. Let's and somebody just jumps on it as soon as possible. [00:07:30] Taylor Kenerson: I I love that. Can you I want you to dive into the element that you just mentioned of, like, this maybe a collaborative effort amongst your team to set the proper expectations for the customers or the people that do come in with those grandiose ideas of what AI could do. And you have to kinda, like, reset the expectations so that you're ensuring that, like, you're both on the same page. So how do you kinda go about that and face a customer or potential customer to to reset those expectations? [00:08:05] Mikus Krams: So if we try to do some kind of Small example and say, hey. Look. You know, this is kinda what you're asking, and we piece together, you know, without properly building it into tool. But this is sort of what you're asking, and this is Kinda how it would look like. And is the result interesting for you or not? And if and usually people are like, it's not really what I expected. And we said, well, You know, we can we can tweak it, but we cannot make it fundamentally different unless, you give us, you know, a lot of your data and then we specifically train it for you. And And we have those discussions as well where people say, actually, you know, I do wanna see what happens when we do that. But, as of yet, we haven't kind of gone all the way down that path, but it, Yeah. It's short short example. Say, look, like, you know, this is this is how it works, when it's At a basic level, and we can improve it marginally, but not fundamentally. [00:09:02] Adil Saleh: Yeah. But it is sometimes it's not possible to, you know, starting off. You don't know exactly what is your ideal customer profile where you you can most penetrate. You don't have a product market for a lot of, a lot of your features, product experiences. They are getting refined by the feedback assistant feedback in the first, like, a year or 2. And, and and I think, you know, we spoke to a lot of companies, and they you know, the mutual kind of feedback that we get in the 1st 2 years is Investing into your onboarding. Invest into your onboarding, and make it as seamless as possible. Maybe, you know, starting off, you can go wide Just like you you mentioned as well, you could you could serve everybody. That's fine to to figure out who's who's the right customer, fit. So, I mean, could you walk us through the that journey, that part of journey? Because, we really we are interested in in in your post sales experiences and how you're driving towards making it standardized and making it more scalable. Of course, 3, 4 years down the road, sitting here, you're selling more to a 150 customers. A lot of them are like, more than 60% of them Our enterprise, mostly of them are, like, enterprise and then mid market as well. So at that point, you'll still be thinking of having not having, like, 15 Customer success or account managers, but you'll be, thinking of having having some sort of scalable model that that, you know, that could at least, you know, standardize 50 or 60% of your your operations postings. So what kind of, you know, process do you have right now? Like, how many customers or maybe Users that that are using your platform, for for for the feedback, sorry, you know, information management and all the, you know, all of that process. [00:10:39] Mikus Krams: So maybe just to quickly step in this playback. So we I think we are going quite narrow. It's it's it's touching, a number of industries, But the persona is very specific. It's called the systems engineer. And so they the the work that they do tends to be quite similar, and it's it's more the the content of the requirements that changes. So but but the actual, let's say, flow is- [00:11:02] Adil Saleh: Their workflow is similar. [00:11:04] Mikus Krams: Yeah. Exactly. And And the way to and actually, we did have a we'd had the very classical, you know, blank space problem. So as you open it up, if it's just empty, it looks like It looks like, an empty Google doc, with some and there's there isn't even a you know, there isn't a tree. There's nothing. Mikus Krams: And so we actually are we're very diligent about what kind of content we have there as you open it up. So the first time you open, we have a sort of a NASA requirements doc, like an open source one. We put it in we said like, this is how it works. There's our own documentation, it's all there. And it's and then you can see how okay, you know, we have these Quality improvement features, and you can you can see them pop up right away. Because otherwise, there's, it's really not possible to It's it's impossible to see like what's the whole point of this thing if it's just empty. But and that's today. And and so as we look towards the future, The the one thing we do wanna work on is there's these, you know, sort of product tours when you open up, it says, you know, click here, do that. I found those to be very helpful. And then even further down, there's these, you know, I borrow from Miro, they have these templates that work really well. And that's that's something that, we wanna build over time is, each of these industries has a some kind of certification that they need to go through. And, like, Not too much if there's ASPICE and and, you know, ISO 26262. And if you open it up and already, you know, you have to be compliant with these, And you have those templates, you know, kind of somewhat baked into the tool, you can get there much, much faster, as opposed to, you know, Typing the first thing and then having to find the find those, sets of rules for the certification then trying to meet them later. It's much much harder. And then, you know, aerospace has something different.  Defense is something else and so on. [00:12:53] Adil Saleh: Mhmm. Yeah. You mentioned, system managers. Now, You know, on the first look, I also realized that it's it's not something that TraceSpace is competing, platforms like Figma, like, just to, you know, Just to talk about your competitive landscape, not talking about, Notion because a lot of these system engineers, they're not even not familiar or they're not using iPhones like Figma or Notion, June. There are so many product management platform that are giving this this kind of, you know, feature set. So they are mostly using, you know, Google Docs maybe, you know, and and they are also collaborative. You can comment. You can have a communication. You can tag people, Invite team members and all. But how do you see, then adopt to the platform, on a on a larger scale? Let's say, I'm not sure how many customers you have or how many system engineers using the platform, but thinking of more than, let's say, 500 in the next 2 years. So how do you see it, differentiating and, you know, all of these folks apart from some of the initiatives you're taking, like product, You know, our beginner guides and, you know, all the, you know, these these startup onboarding experiences and all. [00:14:03] Mikus Krams: So the the The fundamental difference between developing a, software tool and, and some hardware tool is that in software, you only have, let's say, software engineers, it's a single type of engineer and the product is 100% virtual. So your product development tends to be flat. What I mean is, like, there's a single level, and it usually comes from sort of a concept of operations, IE, road map says we need to build, let's say, integration with Jira. And then the, you know, the engineers take that and they they spend maybe a sprint, or maybe I don't know, more or less, but they sort of take it and then they deliver it and then they test it and then they iterate on it. A lot of those things are not possible in hardware because, you know, you can say, hey, I'd love a 4 door luxury SUV, instead of the 2 door one we have right now. But you can't just like, make it and then have iterations on it because each of these iterations costs money. And like even a 3 d printed component. It takes about 6 weeks to be from, let's say, the drawing to getting in your warehouse. Yes. So the prototyping takes a long time. And And and you're also collaborating between very, very different engineers that are not aware of the other engineers, domain. So you know, you think about mechanical, electrical, thermal, Also software and so on. And they all have to work together in a very cohesive way. And so you you almost take this kind of waterfall approach We write out everything and you see how it works in text, then you see how it works in numerical models, then you see how it works in Geometrical model. So like computer aided design, and only then you start really building the, you know, building the car or something. And so it's it's this kind of the multiple layers that you need to have built into your existing product. And that's what we've done. [00:15:54] Mikus Krams: So we have it's sort of a On on the surface, it's relatively easy to document, but each block in the document has a kind of unlimited history. And you can have, a 1,000,000 items and Still no degradation in performance. And and that's that's really makes the difference. And, because Mhmm. Let's say a car could have Maybe a 100000 requirements, and you need to link them all together and they need to work together seamlessly. [00:16:18] Mikus Krams: You need to be able to see if somebody says You know, Elon's book famously has a lot of he talks a lot about requirements. You know, and a lot of it. He's like, you know, why did we have these mats in the car? And then If you can't figure out who's who made you put them in, then you sort of said, okay. We're just gonna remove them. [00:16:38] Mikus Krams: Episode. But that's I think that's the difference in our competitors are really, you know, they come from companies like Siemens and IBM and PTC. So, You know? And then on the other side, you have Notion, and we sort of say, look. Let's bring the Notion, the Figma, the some AI feel into this world, which is dominated by IBM and Siemens. [00:17:07] Adil Saleh: To a company or to a role that is more towards automated automotive and, you know, aerospace, you know, you know, let's say, taking an example of Tesla as you mentioned. You know, from a mechanical standpoint, from an electrical standpoint, from a computer science and AI standpoint, it's It's something that's entirely different than most, you know, most software engineers or, these engineers. They do even system engineers' points. The role can change for you, industry to industry, customer to customer. So how you're managing the bandwidth and, You know, how you're ensuring the skill set and everything. Of course, lot of training and everything. So could you touch us more about that too? [00:17:48] Mikus Krams: Do you mean how do we, adopt the product to [00:17:51] Adil Saleh: different Let's say you have 4 customers, one for auto automotive, Aerospace, manufacturing. What else? Like, defense. So do you think that as well as You do you not think that there's gonna be 4 different skill sets that we need. [00:18:07] Mikus Krams: Yes. But the, fundamentally, the the format of the requirement doesn't change. So they might have You should you should sort of abstract yourself out. The it's just a field of text with some attributes in the history and some links. So fundamentally it doesn't change and and then semi I think if I'm not wrong in semiconductors you don't even call them requirements, you call them items. [00:18:32] Mikus Krams: So Mhmm. But fundamentally, it's the same thing. And the way that, we build a product is to allow to, you know, on an obstruction level, it's all the same, and you can kind of customize it. Do you need a title? Okay. [00:18:44] Mikus Krams: You know, add a title. Do you need a do you need the body? Do you need a rationale. You can have those things. You do need attributes. If, the thing with attributes is if you look at maybe you're familiar how, you know, in Salesforce, You have a lot of fields. Over time, we have like 60 fields. And you can get rid of them. Like every deal has like 60 fields, but you only use like 3. And that's kind of what we've overcome that by allowing imagine in Salesforce, you could have every deal, you just say, okay, for this deal, I'm gonna have x, y, zed fields, but not not the rest of them. So you don't have to configure it at the company level. You can configure it at the At a much lower level. I mean, I don't know. Maybe we're going too deep, but it's Mhmm. Interesting. [00:19:28] Adil Saleh: It's very interesting. Like, I'm glad that, you know You know, we have someone like you on the other stage, and I'm I'm I'm learning a lot onto this. So let's talk a little bit about your your install base. Like, how How many customers do you have? Like, you could talk about users, if not customers, and, give us a story of one customer, as you mentioned, prerecord, like automotive. [00:19:48] Adil Saleh: That's gonna be a a one because I see Bugatti is one of your customer as first. So we'd love to see, like, what what what kind of onboarding, you guys set up for them and, you know, in a on a very high level, how long does it take for a customer in automotive such as to get value out of the platform. Like, do you have any kind of success metrics set up inside any platform that you're using, Salesforce, any CRM, any Data analytics platform. [00:20:18] Mikus Krams: So we we do logging at the back end to kind of track The to track kind of, what what kind of what, you know, how many people how many items are people building? Like, not going into too much details and sort of what specifically they're building. But, the way we start is, we started with a small team because we have to sort of prove the value, within the small team, they, they take some, a relatively new product, or let you know, something that is maybe not the next generation, but a further a bit further out because, and then then It's a separate it's a relatively independent unit that can work on it. It can be either something that's kinda further out For a larger product or if it's sometimes these a single company could be delivering a product every 6 months. Usually when it's sort of a RFP, you know, like some, a larger customer says, hey, we need something. And can you please deliver us 100,000 of these units? And then they and then they they have modules of it, and then they can deliver. And so that's where we would go in either at the beginning of a larger project. Or if If it's a smaller project, it's pretty much anytime. And and usually it starts the way they work. They have to they either have pre existing requirements that they want to reuse, We're just writing new ones, and then we help them we just import them in. Either they can do themselves. There's a there's a import function, but, or we, you know, we also Help them clean it up so it's very useful from day 1. And then they can get started. And depending on what they wanna do, if it's things like we just wanna check the quality, we wanna improve the quality, You can do that right away. You can open a requirement and it highlights you. Hey. This is poor quality. You need to rewrite this. If it's, more about collaboration, then, of course, The value is shown over time where, for example, from one of like, something that takes us acute you know, a couple of clicks to it. You know, one of our let's say, the largest competitor, it takes literally 15 seconds every time. And if you have to do this 200 times a day, then it just adds up and you get really annoyed. Episode. So that's what's tedious. Yeah. [00:22:31] Mikus Krams: Yeah. It's very tedious. And so that's the, You know, I mean, whether you consider that value or not, but I think it adds up and then it literally saves you not only time, but it also saves you Nerves and brainpower. You know? [00:22:45] Adil Saleh: Yeah. Even let's say, taking example of, Bugatti Day, let's say, a year back, they have launched a model. They want more units of those models, like more, mechanical, 3 d modeling, whatever. So a lot of them that has been already predefined by them, but they might not be, Taking some quality short measurement up to the mark and, you know, you're helping them optimize the time, quality, all of that. That's also a big use case in in quality manufacturing. And in cases that they need more newer models, when they're launching a model, that's that becomes a different different case. [00:23:18] Mikus Krams: Yeah. The example I gave about sort of having a lot of iterations, it's more about, companies that produce components for someone else Rather than a car because a car would probably take a few years, you know, from really the beginning to all the way to the end. But, some components in cars, it can be some power units or some some some stuff, you know, That, Mhmm. And that can that can be very, very quick. You know? Somebody says, you know, please deliver a 100,000 of these or, you know, a1000000 of these in 6 months, then To spec, then you have to kind of Mhmm. Take their spec, develop your own, understand if you can actually deliver it, then build it and deliver it. Episode. It's very, very good. [00:24:01] Adil Saleh: Mhmm. Quick last question. We are pretty much on time. Now how do you see the product stickiness? Like, platform stickiness to these Customers mostly in automotive and, car manufacturing, aerospace, all these categories that you're trying to penetrate. Like, it is it like not Is it, like, not the same like most of these, companies that are using data analytics platform once they've migrated all of their data inside? I mean, they, of course, they make sure they do the due due diligence and everything. But once they're inside, it's so hard for them to, you know, you know, quit on on a on a platform. So How does the product stickiness platform stickiness look like? [00:24:37] Mikus Krams: Yeah. It is. It's a very, it's a good question because it is a a relatively big challenge. But I liked with the Jason Lemkin, I think, once wrote about, that in in a year's time frame, the it might be quite hard to unseat an existing player. But if you look over 3 years, there's usually enough time. [00:24:56] Mikus Krams: Like, they kind of if they don't like the existing product, they will churn over the 3 year time because they might have a 2 year contract and they might take a year or 2 off board. But in a longer timeframe, it does work. During our stage, we're we're trying not to go, you know, fully head on, especially if people have recently migrated, Because it's not just the cost of the tool. It's it's not just the quality, but it's also literally they've just put in, you know, let's say, 9 months of work migrating. They're unlikely to want to take the call of saying Yeah. Let's do another one of those, you know all the stuff all the work that you did is for nothing. So we we usually tackle there's a lot of, kind of workflows where people, they have chosen not to use a tool because in the past they've for experience. And that's where we go in or they are aware of them and they haven't, picked one yet because as well, you know, their their mom, episode. It's not what they expect, and it's maybe more expensive than they expect. So it's these kind of reasons that, where we say, look, look, here's a better offering. [00:25:56] Mikus Krams: It's a differentiator. It's, episode. It's it's more modern, so you can come in and start using it right away. So the onboarding for us I mean, it's sometimes it's ridiculous. Some of our competitors, it takes, like, a week or so just to get started, like, literally to just get started. And it's an unimaginable in something like Notion or Figma. Right? So for us, it's the same thing. You can just open it up and get started. And And when we do demos, we literally say, hey. Here's access. Give it a shot yourself. Like, did you expect you would be able to onboard yourself in an hour? Probably not. Episode. Here we are. [00:26:29] Adil Saleh: That's the way to go. That's the way to go to make sure you keep your platform as self serve as possible from day 1. Episode people come up with with questions, and then you build some sort of beginner guys, as you mentioned, that, you know, videos and all of those. That's good. Okay. It was real nice meeting you and getting to know, face space, and it was so interesting of a conversation. You know, I get to learn a lot personally, And, I wish you good luck for the year 2024. I'm I'm I'm so sure that you'll you'll have some big car manufacturing. That is only use case that I understood during this conversation, but, manufacturing companies, using, using your platform. [00:27:07] Mikus Krams: Yeah. Thank you much. Thanks for having me and good luck and happy new year to both. [00:27:11] Taylor Kenerson: Yes. Thank you so much. Have a beautiful night. We'll talk soon.

Keep Listening VIEW ALL EPISODES >>>