In the world of infrastructure, underground pipes don’t exactly get a lot of love. Hidden from view, they silently transport water, sewage, and storm runoff, serving as the backbone of our urban landscapes. But when things go wrong down there, the results can be catastrophic—think burst pipes, sinkholes, and flooded basements.

That’s the kind of trouble, VAPAR aims to fight, by leveraging the power of Machine Learning and sound commitment! How? Let’s find out:

with 🎙️ Amanda Siqueira – Co-Founder & CEO of VAPAR

---

Resources:

🔗 VAPAR’s website

🔗 Peter Thiel’s Zero to One

🔗 Everyone Hates Marketers on Jobs to be Done

🔗 David Lloyd Owen’s appearance on the podcast

---

Full Video:

---

Introducing: VAPAR

Until recently, inspecting these pipes was a labor-intensive, time-consuming task often left until a problem was impossible to ignore. Enter VAPAR, a game-changer in the field of underground pipe inspections. Leveraging AI technology, VAPAR makes inspecting these pipes not only seamless but smarter. Let’s dig into how.

The Problem with Traditional Pipe Inspections

Picture this: an engineer sitting for eight hours a day, glued to footage of dark, murky underground pipes. They’re meticulously scanning for cracks, blockages, or any signs of trouble. Now imagine this process being repeated across thousands of miles of pipes—tedious, error-prone, and expensive. Manual inspections require operators to watch the same footage multiple times, increasing the likelihood of missed defects and inconsistent results. Moreover, the whole ordeal is incredibly costly, with some inspections costing between $5 to $10 per meter, while repairs can skyrocket to $1,000 per meter or more if damage is found​.

But the real kicker? Even with all this effort, most utilities still operate reactively, waiting for customers to report issues before inspecting their infrastructure. This outdated method doesn’t just waste money; it risks lives and property by allowing small problems to become big ones.

Enter VAPAR: The AI-Powered Solution

VAPAR is turning this process on its head by automating pipe inspections with artificial intelligence. Amanda Siqueira, co-founder of VAPAR, knows the pain of manual inspection all too well, having once spent her days watching sewer footage at Sydney Water. Along with her co-founder, Michelle Aguilar, she decided it was high time to take a more proactive approach to pipe management.

With VAPAR’s AI, the inspection process has evolved from hours or days of footage review to mere minutes. VAPAR’s AI-powered platform takes raw footage from cameras inserted into underground pipes and instantly identifies potential defects like cracks, root intrusions, or blockages. Instead of relying on multiple manual reviews to spot these issues, the AI does it in one go, with greater accuracy and consistency​​.

From Weeks to Minutes: A Seamless Transition

Think about it: a process that once took weeks or even months now gets done in minutes. VAPAR’s platform has automated the entire inspection workflow, from uploading footage to identifying issues and even prioritizing which defects need immediate attention. It’s like having a super-efficient assistant who never gets tired or distracted and doesn’t charge overtime.

The speed and efficiency are not just convenient—they’re cost-effective. The AI detects more defects than human eyes typically would, without the fatigue that comes from staring at footage for hours on end. This means more problems are caught earlier, and fewer expensive surprises crop up down the road. By automating defect identification, VAPAR helps utilities cut inspection costs by up to 90%​​.

AI is a Partner, not a Replacement

You might think that with such high-tech automation, VAPAR’s AI would replace human inspectors entirely, but that’s not the case. VAPAR views its AI as a “co-pilot,” working alongside human operators to make their jobs easier, not redundant. While the AI takes care of the grunt work—scanning footage and flagging potential issues—it still leaves final decisions to experienced engineers. This partnership approach ensures that the AI’s findings are always reviewed and validated by a human, adding an extra layer of quality control.

VAPAR also emphasizes transparency. The AI isn’t a mysterious black box; it provides clear, interpretable results that engineers can understand and verify. In practice, this collaborative model means that even though the AI flags more defects, utilities don’t need to panic. They can apply their own risk assessments to decide which issues need immediate action and which can wait​.

Real-World Success: The Northumbrian Water Case Study

One of VAPAR’s recent success stories involves Northumbrian Water, a utility company in the UK. Traditionally, Northumbrian Water operated a reactive CCTV program—only inspecting pipes when a customer complaint was lodged. But with VAPAR’s AI, they were able to streamline their inspection process significantly. Footage was uploaded directly to VAPAR’s platform, where the AI analyzed it for defects. Northumbrian Water’s technical specialists then reviewed the AI’s findings, turning what was once a highly subjective and manual process into a quick, data-driven decision-making exercise.

The result? Not only did this save time, but it also aligned maintenance activities more closely with asset health, ensuring that repairs were targeted where they were needed most, and preventing unnecessary work. This has allowed Northumbrian Water to prioritize their investments better and reduce the risk of catastrophic failures​​.

VAPAR’s AI is Proactive, Not Reactive

So, what’s the big deal about VAPAR’s AI-powered approach? It all comes down to being proactive rather than reactive. Most utilities are caught in a vicious cycle of responding to problems only after they arise, wasting precious time and resources. VAPAR flips the script by using data to predict where issues are likely to occur, helping utilities to get ahead of problems before they escalate.

By transforming inspection data into actionable insights, VAPAR enables utilities to plan their investments more strategically, ensuring money is spent where it matters most. In an industry where budgets are tight and stakes are high, this kind of efficiency isn’t just a luxury—it’s a necessity​​.

A Broader Vision: Beyond Pipes

But VAPAR isn’t stopping at just underground pipes. Their vision is broader—they’re eyeing all public assets that require regular inspection and maintenance. From roads to bridges to water treatment facilities, the potential applications for their AI technology are vast. As infrastructure ages globally, the need for smarter, more efficient management solutions like VAPAR’s will only grow​.

On the Horizon

By combining deep learning with real-world expertise, VAPAR is reshaping how we think about infrastructure management. Their AI-powered platform offers a seamless, efficient, and proactive approach to underground pipe inspections, reducing costs and preventing failures before they happen. And while they might be starting with pipes, their ambitions hint at a future where the same smart, streamlined approach could be applied to all kinds of public assets.

In a world where we expect everything to be instant and effortless, isn’t it about time our infrastructure inspections caught up?

---

My Full Conversation with Amanda Siqueira on VAPAR

These are computer-generated, so expect some typos 🙂

Antoine Walter: Hi, Amanda, welcome to the show.

Amanda Siqueira Legacy: Hey, Angela. Andthanks for having me. I’m really excited to be

here.

Antoine Walter: It’s been a while. I had an Australian series at some point with a lot of guests from down under. And it’s always a pleasure to think that, you know, we have all those thousands of kilometers who set us apart. And on your website, actually, there’s a big number Featuring those millions of kilometers because you’re saying that you’re helping 6 million feet of pipes. to be managed. And I’m just wondering what is the status quo you’re replacing?

How would their life be if VAPAR wasn’t to exist?

The Problem with Manual Pipe Inspections

Amanda Siqueira: We’re expanding our footprint, which we’re very excited about. So, last year we did over 6 million feet of inspection footage and helped with the investment planning of that. Without. VAPAR, that whole process would have been painfully manual. people would have had to, much like myself when I was an engineering intern, gosh, now many years ago watching eight hours a day of footage and just writing down whenever I saw anything weird inspection footage of an underground pipe.

All of that process is now completely automated in our platform using deep learning to identify defects and things like that. So tens of process that could be, weeks, sometimes months with some of our clients into a couple of minutes. Very customizable, very easy for them to come to a decision as to where they

spend

their money.

How VAPAR Automates Pipe Inspections

Amanda Siqueira: So that’s footage of wastewater and stormwater lines, Yes, exactly. So they’ll put cameras inside sewer pipes and stormwater pipes just to try and find out, is there any blockages? Are there any breaks? Trying to get him ahead of the problem so that, you know, people don’t just fall into a sinkhole or have sewage come up in their basement.

So they’ll try and put these cameras in and try and find the issues. But all of that footage.

every foot, every meter has to be watched manually. Very often, more than once. So because the process is manual, it’s error prone, , if you watch it, I’m not gonna believe you, or if I watch it, you’re not gonna believe me.

So we’ll watch it twice which is insane. Given that over , a billion meters is inspected every

year.

Yeah.

Antoine Walter: go into who’s your customer their life, how they integrate with what you do and what you exactly do. But before that if I try to boil it down to what’s their unique value they’re getting from you, because I could see it as. Time saver as a money saver on the long run, on the short run, what is like this absolute spice where they realize that with VAPAR there’s this.

Amanda Siqueira: basically Our whole company and our, all our product is focused at doing is fixing pipes before they fail. So most of the world runs their pipe inspection and management process reactively, unfortunately. They’ll wait until a customer rings up and tell them that there’s something in their front yard or in their basement and then they’ll put a camera in.

With our product, we’re basically helping them use that data to try and run a more proactive program by helping them say, well, this is the condition of your asset. These are the highest risk assets. You should go target investment here. And the next time you either don’t or do need to come back to this spot.

B

Understanding the Impact of Pipe Failures

Antoine Walter: ut how bad is it if it breaks? Because on the waterline, I get it. It breaks, you have a main damage, probably the roads gets damaged. I saw some statistics from Canada. a while ago, but they should pertain true, which is about count 1 million of impact for one break on the main waterline, but you are on the sewer sides and on the stormwater side.

So is it similar impact we’re discussing here? Or what’s the harm

Amanda Siqueira: gravity networks, like stormwater and sewer, sometimes thought of as the poor cousins of drinking water. They fall under gravity. They’re not pressurized like drinking water. And the consequence of failure is slightly different.

some might say a little bit lower because it’s not related toa human right. The way that drinking water is but the consequence of failure sometimes can be catastrophic. especially when you get your large trunk mains the condition of that is super important.

If they fail you could get large sinkholes. So there are many instances across the world where sinkholes form and people Unfortunately driving to them or all the things that, you know, consume houses even but there’s also the environmental impact of that. So not even, you know, human health and public health reasons, but even just environmental impact and sewage in the waterways you only have to do a quick Google of the UK public media at the moment to see how that’s being reflected and played out at the moment.

Antoine Walter: just before we go into the deep dive, I have a sanity check to do, which is you’re making. CCTV inspection better because you’re removing the human aspects of it. Like the human subjectivity, it becomes objective parameters checked by AI. I get that. My sanity check is on the part of how many of the, of the Thousands of kilometers of stormwater and sewer are already inspected with CCTV. Because if you’re making CCTV better, but CCTV is , like really a small portion of the inspection, then you’re making a portion of the portion better. So what’s the status quo?

The Global State of Pipe Inspections

Amanda Siqueira: like I mentioned before, over a billion meters of pipe is inspected every year. And that’s roughly 10 percent of the global network. But the thing is that global network is inspected every year. So that billion meters is recycled every year. The attempt of most small utilities and certainly the ambition of most small utilities is to try and see if they can inspect their entire network at, , a one to 10 frequency.

So they get around to the same pipe every 10 years most. Water utilities are not anywhere close to that. Some are between, you know, one and 20 years. Like we see in the US in other markets such as Australia and New Zealand and even the UK, it can go out to one every 50 years, once every 80 years.

But crucially, they’ll keep inspecting their network Until they reach that frequency. So, even though it’s a discrete amount every year, they’ll do the same and try and get that number up as much as they can to bring that frequency down.

Antoine Walter: Let me try to see if I get that one. So there are about 10 billion meters of pipes, which could be inspected out of which every year, 1 billion gets inspected. You’re inspecting 2000 kilometers. So 0. 2 percent of the amount, which is CCTV inspected. And the added value is that today they might just inspect, and see things which maybe evolve.

So they need to do it every year and stuff like that, where you say, I’m assessing, and then you’re good to go for a longer period of time. Do I get that one? Right?

Amanda Siqueira: Yes. Yeah, exactly. And there’s huge upside. So like you say, VAPAR’s adoption is a small percentage of the global inspection market. So there’s a few jobs on what we do, but also, it’s a lot of money to put a camera in a pipe.

So if you’re going to spend the money to put the camera in the pipe, you want that data to be high quality. You want it to be objective and. Very very accurate. So that’s what we help with as well.

Antoine Walter: Interesting. It’s a lot of money to put a camera in a pipe. How much additional costs would it be to run your inspection on top? If you take a percentage, like it’s 100 points, put that camera in the pipe, how many points would it be for you to inspect on top of that?

Amanda Siqueira: We’re sub 10 percent of the meter per meter cost of putting a camera in the pipe. But that’s not actually where our business case for our software is actually made. Business case for our software is what happens after the camera comes out of that pipe. So someone’s going to make a decision as to whether I need to do anything or there’s any activities that need to be scheduled.

And that?

can bethree to eight times more expensive, that activity. And so that’s the decision that needs to have a little bit more governance over it based on the data that’s collected.

Antoine Walter: I get that it’s not the business case and we’ll discuss the business case, but I’m just trying to get, a muggled understanding of all of that, because if I take a reference of when you’re welding pipes, for instance you have some norms, which tell you, you should do destructive testing.

And then destructive testing is super expensive because you need to cut your pipe, cut it out see if the world was right or not and stuff like that, which means you have 100 percent more costs when you’re doing that additional layer. Compared to if you did nothing, whereas here, if I get you right, you have less than 10 percent additional costs to run the VAPAR on top of what you’re already doing, which might make you think, okay, it’s not that much more probably worth the effort, but we’ll go into that.

The Founders’ Journey

Antoine Walter: Before we go into the absolute details, I’d like to get your story right. Ends. You have a slightly different story on your website from what I read from press articles, because you are two co founders with Michel Aguilar, who’s your co founder.

And your website says two Australian engineers on a mission. Whereas what I read in the press is that you’re also high school friends, which is not mutually exclusive, but what’s the story?

Amanda Siqueira: Michelle and I went to uni together. , so I did civil environmental engineering and Michelle did mechatronics engineering. So all things machine learning so knowing each other from uni we also went to high school. You’re right. We went to, Western Sydney high school together.

So we’ve known each other before we started VAPAR for over 15 years and then much longer now since we’ve started VAPAR together.

Antoine Walter: You know, each other and you go on to do your own engineering studies. And within those studies, you end up having to inspect pipes and you find out it’s cumbersome. And then you decide to do something about it. It sounds like the perfect Silicon Valley story, but there’s always more nitty gritty details to it.

So what’s the full story?

Early Challenges and First Customer

Amanda Siqueira: Yeah, so I was working at Sydney water as an intern eight hours a day, just watching sewer footage. But that was in 2009 when I was an intern and then VAPAR only started in 2017, 2018. Michelle herself actually had a bit of wastewater experience as well. She worked on the process automation controls for Bondi wastewater treatment plant.

So famous beach Bondi the wastewater treatment plant controls on that. So she had a little bit of experience with not only Sydney water, but with um, Um, and then, , her and I have always been tossing ideas back and forth with each other. So when, deep learning a lot of the open source frameworks were coming out in 2017 and they were starting to beNot only the frameworks itself, but the forums were guiding people on how to use these open source frameworks and people like Michelle and I could just jump on the forums and learn really quickly how to use these frameworks and customize them to our very specific needs.

Not everyone knows what they’re looking at when they’re looking at inspection footage. So, at that time it was a really nascent spot in the deep learning field. So, yeah, we were able to capitalize on it really

early.

Antoine Walter: You say forums. Did you teach yourself deep learning on Reddit or what’s the story?

Amanda Siqueira: Oh, you’d be surprised. Well, I mean, at that time there was the forums was one of the only ways. I mean, short of that, you had to kind of be a PhD in computer science to be able to like replicate what Google was just giving you for free.

Antoine Walter: But still, there’s a word between we always wanted to do something together. We noticed that field was maybe under researched, underserved, and we create a company and we take the world. So at what’s the turning point? When do you say, Oh, we need to create that thing and that thing has to be VAPAR.

And it’s going to be that, was it really what we see today of VAPAR or did you do some pivots along the way? What was the first iteration?

Amanda Siqueira: Yeah. So, the first iteration of VAPAR was really, again, capitalizing on that deep learning object detection kind of movement that was really early in, in 2017. And it was basically AI for anything. So it was, you know, counting people, counting cars, counting traffic lights, counting possums.

there was lots of kind of use cases that we were exploring and trying to test, you know, will people , pay us for this sort of data. then I want to say it was a book by Peter Thiel zero to one that I read and there was a whole bunch of criteria that he was kind of, you know, he puts out there to say what makes a good startup and especially in that kind of early phase.

And one of the things that I think really crystallized. The VAPAR that we know today is what is something that you as a founder team like what’s your secret sauce that like no one else or it would be very difficult for someone to replicate. And, you know, Michelle and I kind of looked at each other and. There’s myself and just as a rough number, some 3, 000, some 5, 000 people globally that know what they’re looking at when they’re looking at the inside of a sewer pipe. I hope that number will change, but at the moment, very few people know what they’re looking at. And so if Google wanted to try and replicate this, it would be very difficult now, maybe not the case, but when we were starting there was much more interesting things that Google were applying theiralgorithms towe were like, this is a small space would probably protected for now.

And so we decided to start there and lo and behold, people started paying us for it. and certainly we built it out from there.

Antoine Walter: That’s super interesting and I’ll continue with that one but just, sorry I’m slowing you down. But it makes me think, you know, of when chat GPT-3, 3.54 came out, you had a lot of entrepreneurs coming out and saying, Hey, I can now do that with chat GPT, which I could not do before. I’m putting the brackets around entrepreneur, because a lot of that was like, really when the new chat GPT was out, their business model was dead. What you’re saying is that. That was the mass market reaction to AI becoming main stage. You were an early adopter in the crew of people who in the 2017, so those algorithms were starting to be public and to be in the open space with those forums. And then you’re thinking, okay, that was early adopters. What is special about us? We understand the shitty side of the story, so let’s go there. But still you had to have some merits to be within those early adopters, which makes me think you were on a chase to create something. Am I right with that?

Amanda Siqueira: Yes. Yeah, definitely. I think even me personally, really fell in love with the idea of starting a business, starting something. , I did engineering um, thought that?

I needed to get an MBA. So I spent about a year trying to get into an MBA. And another engineering colleague of mine who had started his own business, he was you just got to start, there’s no MBA in the world that’s going to teach you anything.

you won’t learn quicker if you just do it yourself.

Antoine Walter: You mentioned that not long after you had create your first MVP, some money started flowing. So what’s your first customer? What’s the story?

Amanda Siqueira: Yeah. So our first customer was a council in Sydney, Australia they had a huge amount of inspection footage 5, 000 videos, in fact that they had never got coded but they wanted to be able to pull into their system and start making investment decisions on and.

It was kind of like this serendipitous moment where we were like, well, we need footage to train our algorithm. And so Michelle built this very early stage of our product where we actually went through and manually coded the footage for them.

So it was more of a professional services arrangement. I wasn’t even a product really at that point. By delivering that project we were developing that data set as I went. So, yeah, it was a very interesting project. We spoke openly about the AI ambition with them, and that won them over because, you know, they could have got that done by anybody.

But they wanted to get us involved and we were very excited about being involved because, they saw the vision to say, this is ridiculous, that we’re getting people to do this. Like, if you can find a solution, go

for it.

the first AI was you. Yeah. Well, you’ve gotta start somewhere. But now there’s a lot more ed to the process.

Antoine Walter: do you find them? Do they find you? What’s the way

Amanda Siqueira: certainly At that stage, it was just I had worked in the industry. Michelle as well had worked in the industry. It was really just networks. So people that we knew, contacts that we knew that sort of stuff. I mean, even now It’s a lot about warm introductions and referrals, clients speaking to clients.

So that’s never really gone away. I think the wood industry, as you know, is very small but tight knit. So yeah, that’s been helpful.

AI in Pipe Inspections: Benefits and Challenges

Antoine Walter: When you’re looking at those footage images, what are you looking for? Are you looking for, to say, you know, that’s a route, that’s a crack, that’s whatever could be like a fatberg or I don’t know what could all be in those pipes? Or are you also straight from the beginning assessing that and saying, yeah, that’s a route.

You’re right. But given how the route is placed and don’t touch it for the next 10 years

Amanda Siqueira: It’s a bit of an iterative process. , Each region that we’re in. So we’re in Australia, New Zealand us and the UK there is a reporting standard for each of these. So if you’re doing pipe coding there’s a standard against which you need to do that.

And there’s on average, roughly around 200 particular defects that a reporter could report on. And so our product or even a manual process would be going through and looking at the footage and saying, okay, If I say something that’s interesting does it classify against any of these 200 defects that are in the standard?

So that process happens first and objectively and rightly so, the layering of this condition assessment or risk assessment generally happens after To say, okay, well, I found all these defects. Maybe these ones I care about these ones. I don’t. The severity is also taking into account things like that to make sure that you’ve got all the granularity and you’ve got all the information.

So that, you know, if you inspect it in 10 years, again, you can see the, potential depreciation of the pipe, but that decision making that point of time generally is separate from the actual defect identification itself,

Antoine Walter: As a rule of thumb, AI detect more defects than the human eye or less defect than human eye or exactly the same number

Amanda Siqueira: more actually and we found that phenomena that The AI detects more because there’s not as much fatigue associated with identifying like fine roots at every joint. you need to report on any root intrusion. in clay pipes, there’s a joint every two meters.

And so that can be cumbersome. And if you’re an operator, you have to like basically pause and write physically fine routes into the inspection software. So, AI picks up more. But when it comes to severity operators basically it’s par for par if it’s high severity, so they’re very unlikely to miss human operator would very unlikely to miss things like collapses or big holes in the pipe as

same with AI.

Antoine Walter: but the fact that it detects more is what I was expecting. And where my question is leading to is can it be scary for the utility to suddenly see that they are much more defects than what they thought because you’re detecting more and hence they might be overreacting. Whereas what you’re detecting is something which is of a low severity, which is good to know, but also is proof that.

After all the network is safe

Amanda Siqueira: Yes. . It can be very scary and a lot of utilities for that reason, kind of. Don’t like the granularity that the data that the AI can give you which is totally understandable because as soon as you write these things down, I’m an organization says you write these things down and they have to, act on it.

But the way that we work with our clients is to say, Hey, this is the data. This is what it says. But your risk assessment and your asset management policy can dictate something different. So just because you have all of these defects in the network we can have a look at them, apply a certain risk score to each of these defects and say, this is something that is going to impact in this financial year or this planning period or not.

And we can start to calibrate how the treatment is applied so that their longterm financial planning is not. completely blown out of the park just because they found like 15 cracks that they’d never

seen

before.

Antoine Walter: and do they trust you?

Amanda Siqueira: So that’s a good question. We’ve shown in our data that our AI has been able to pick up 90 percent of the defectsin people’s inspection footage. So there is a percentage where people have had to go and get involved and Nope, the AI is wrong. But we make that completely open and available for people to do that.

So within the product itself, they can see what the AI is thinking. We don’t make it a black box for them. they can say, okay, AI has got this wrong. I can see why it’s got it wrong. So we show people what the AI is thinking. we try to engender a sense of trust, but make it very customizable so that they’re making the decisions.

And that’s, I think, a really important part because we want to empower our clients to understand impacts of their decisions.

Antoine Walter: I didn’t ask that to push in a corner. It was just, you know, I’m working for a 222 years old piping company. And we introduced, I think a bit less than a decade ago, a non destructive testing of pipe welds service. And our first customer, our value proposition for that was to say, we inspect the welds,

 And we’re going to assess, and we might sometimes find defects, but we’re going to tell you those defects are fully fine, are within the design parameters, hence keep the weld, don’t cut it out. And we still think is the full added value of the service, which is to say, you don’t simply see a defect and cut out instead of what you do the maths.

And if the maths add up that you still have sufficient strength in the world, you would not touch the world. But that customer told us, yeah, all nice. I read your story. I listened to your marketing. I read the brochure. I saw the video, whatever you want, but if you find a defect, whatever it is, cut it out. defeating the point, but it’s a matter of trust. And I’m thinking if we come as a two centuries old company and we give an assessment, we give a guarantee and still the customer wants us to cut out, just to say, yeah, machine can calculate stuff, but I’m running my plant, you know, we did like that for 10 years or for 20 years or for 40 years, which you.Probably here pretty often. I’m thinking as a young startup coming out fresh in the field. And you’re saying, Hey, we detected all of those, but it’s fine. It’s not that straightforward that they trust you from the get go

Amanda Siqueira: Yes, you’re right. As far as their risk assessment we basically say to them, these are all the things that we found. And then we also go back to them and say, okay, what’s your thoughts on cracking or what’s your thoughts on root intrusion?

Do you want zero root intrusion? In which case you got to dig up every pipe. Or are you saying to us actually 10 percent of a root intrusion of the cross sectional area of the pipe? I’m okay with that. Because I inspect frequently or I’ve got a return program already online , that calibrates down the investment that’s needed.

So it’s a really, yeah we totally involve the client to make sure that they’re they’re making the decisions, but you’ll rewrite some clients just have a blanket zero defect policy, which is very challenging if you’re dealing with taxpayer dollars, but nonetheless, some people have the money.

Antoine Walter: So for all your potential customers who might be listening to us, let’s take. one of your projects from A to Z, where you’re going to tell me who’s your ideal customer, how do they look like, and , if one of your ideal customers happens to really be your customer, let’s pick that story.

I know you have stories from Australia, from New Zealand, from UK, even one from the US. So I let you pick the one which you think would be the most accurate to describe your process. how you initiate the conversation with them, what happens, what do you deliver, and when does it end, if it ends.

So, what story do we tell? What case study do you want to pick?

Amanda Siqueira: Yes. Yeah.

Case Study: Northumbrian Water

Amanda Siqueira: So it was actually a recent case study that we delivered at the swan conference in Vancouver a couple of months ago. So we do have a case study with Northumbrian water. They’re a water utility in the UK. Albeit the UK is a very Unique market, but the use case that Northumbrian water used us for isn’t so .

They run a reactive CCTV program. And what that means is basically the customer calls up and tells them that something’s gone wrong before they put a camera in the pipe, as opposed to proactive, where they putting a camera in the pipe before the customer calls up. They don’t want to do that, but they’re forced to, because there’s a lot on their network that they’ve got to manage.

And so they came to us actually via another customer referral. it was another customer United utilities that was advocating the use of us and Northumbria had found out about us through them They actually did a trialwhere they pitted us against another AI company and selected us because of, various parameters, AI accuracy product ease, things like that customer serviceWhat we did for Northumbrian Water is we took all of their reactiveCCTV inspection footage and people uploaded the footage directly from site.

So there was no coding on site. They uploaded the footage directly to our platform. The AI did all of the coding and then the people in the back office at Northumbrian, the technical specialists, they came in and said, Based on the defects identified and the recommendation that we had applied they decide, do they agree with the recommendation or the AI or do they not?

And so it turned this process that is very subjective into a very streamlined process. But the other thing that it did that really helped Northumbrian Water was aligned asset health with investment. So the crucial thing there is when you’re putting a camera in a pipe to fixa blockage or something like that the asset management principles of the organizationmay not be front of mind of the person that’s going and popping that blockage.

And so to be able to get data to match up to say, okay, this blockage is because there’s a joint displacement. And so I need to fix that joint displacement so I don’t come back here. Again be identified in the back office and that process is then eliminated. So they don’t go back again.

Antoine Walter: Lots of stuff to deconstruct here. So let me take them one by one. First, if I get you, right, that means with Northumbria and what you did was automation level three. So you’re the footage and giving a recommendation as to what should happen, and then someone manually steps in and reviews that.

I don’t know if it’s 100 percent of them or if they were sampling some of them then that I agree with the recommendation that I don’t agree with the recommendation. So is that your streamlined way of working or was it a way to build the confidence with Northumbria? And now they would simply trust you because they found out that they would agree with the recommendation 99 percent of the time.

Amanda Siqueira: No, so that’s the way that we work with all clients. So even though we strongly advocate the use of AI, the way that we talk about our tool is it’s like an assistant, you know, to borrow the Microsoft term of copilot. It’s like an assistant. It’s like an extension of your team.

It’s like a grad or an intern in your team. So you let them help you or certainly they can be helpful. But, You wouldn’t leave them unsupervised. So all the footage has a recommendation against it, but all the footage does need to have someone to go through and say, yes, no, yes, no. And that’s the thing.

It’s exception handling. It’s not, you know, the detailed analysis. And so that’s the offset here is to say, okay, you don’t have to do the detailed analysis but you do have to do the exception handling, which is much quicker.

Antoine Walter: but do you have like a statistics of how often the human would disagree with the AI?

Amanda Siqueira: 10 percent of the time there’s a difference of opinion.

Antoine Walter: in the long run, would you envision a future where that validation step from the human falls out?

Amanda Siqueira: I think we’ve definitely toyed with the idea but , what I expect at the very least is that level of human intervention to go down because we know that the volumes of inspection footage is going to go up. I don’t want to be enabling people to just sit there and do exception handling the whole time.

But to remove humans from the process altogether I think would be a mistake. I think for the base needs of skill preservation, but I think there’s also justwe don’t want to be removing humans from a process. And I think it’s a strong case for enablement but not

replacement.

Antoine Walter: Are you saying that because you believe in it? Or are you saying that because you don’t want to be too scary with your AI stuff?

Amanda Siqueira: No, I genuinely believe it. Like, so even if you look at mypersonal use case, and this is a lot of the reason of my personal why behind VAPAR, I was watching footage eight hours a day is my job. And it wasn’t the watching the footage that was exciting and certainly nothing about watching footage is exciting.

But the more complex stuff is like, what do I need to do about it? Do I need to do it now? How much is it going to cost me? You know, how can I dovetail this with maybe roadworks that are also planned? So there’s a huge amount of complexity that goes into the decision making that engineers and technical people do need to be involved in.

And it’s not a straightforward set and forget that you can just leave it to. So yeahit’s about removing the mundane stuff and, you know, kind of helping get to the more complex stuff that people do need to be involved in.

Antoine Walter: you said you expect to see more of that inspection to happen in the future and more of it to be checked. What is the bottleneck today? Like, could you be having robots, like the Mars robots, just wandering the sewers day in, day out? And the reason why we don’t do that is that today it needs humans to watch it.

And if they don’t watch it, you’re just capturing data for nothing. In which case you’re somewhat removing the bottleneck because you’re simplifying the inclusion of humans, or is it still prohibitively expensive to run those robots in the sewer? In which case it might grow as a practice to inspect more, but not spectacularly.

So on which end would we be

Amanda Siqueira: it’s on the ladder. So it is prohibitively expensive toinspect your entire network. And even though that’s the start of the value chain, we’ll get more and more footage as there’s more and more footage collected. The value proposition of what we provide is more the data insights that comes from , the aftermath.

And so even though there’s mediocre growth in how much inspection footage is captured every year we’re just chasing our tails on that number.

Investment in Water Infrastructure

Amanda Siqueira: We’re going backwards. There’s billions of dollars that are being spent every year.

And that’s. Not even close to what needs to be spent. So, the investment side of things is going up and needs to go up further.

Antoine Walter: There’s a book, which is difficult to read. Not the zero to one, like very good recommendation you gave us. It’s really you have to be a water nerd to find any interested in, but it’s full of wisdom stats and stuff like that, which is called Global Water Funding, which is a book by David Lloyd Owen.

And I had David on that microphone some seasons ago to discuss the book, but his conclusion is to say that we have to do more with less just because the infrastructure investments are not going up to the magnitude it should, if you want to keep just the networks in the status quo they are today, not even thinking of making them better.

So, We have to be much more cautious about the investments and the way we place them so that we make them more efficient, which makes me think that exactly what you said, when you avoid replacing a line, which doesn’t have to be replaced. And when you strategically pick the one which has to be replaced at the right time and for the right reasons, well, you’re doing exactly that you’re doing more with less or same budgets.

You mentioned these three to eight times lever aspects between. the cost itself of the inspection and what it really delivers in terms of infrastructure management. First, let’s get that one right. What are we exactly discussing here in terms of infrastructure planning? Can you please Qualify those three to eight, and then we go into your scope.

Amanda Siqueira: Yeah.

Cost and Efficiency of Pipe Inspections

Amanda Siqueira: so let’s take Australia, for example, it can cost anywhere between, five to 10 a meter to put a camera in a park. But then it can cost between, 250 to 1, 000 a meter to do any sort of rehabilitation on that pipe based on the inspection footage that was captured. And that’s only if something needs to be done.

So obviously not all the time, sort of one to one ratio of inspection to rehab. But if something does need to be done, it’s very expensive to do it. Whether it’s trenchless or whether you have to dig the trench. Very expensive to do the work.

Decision-Making in Infrastructure Management

Amanda Siqueira: So, that’s where the decision making is super important to make sure that you’re only spending that money on things that are high risk and deescalating even Assets that you are promoting to rehab and saying, actually, maybe I don’t need to do that this financial year.

But I have budget potentially later on.

Antoine Walter: So five to $10 per meter to inspect 250 to $300 per meter to rehab if needed. Out of these places where you expect to see trouble, what’s the rough rule of thumb?

Percentage of pipes where you’d say. You need to do something in the coming five years.

Amanda Siqueira: Yeah. So we find about 10 percent of the inspected network requires some sort ofrehabilitation or follow on work. And that can vary. Like you say, sometimes there’s a program that’s actively targeting an area where they know that there’s issues and some of it’s proactive.

So it can go up and down from that.

Antoine Walter: But if, I do quick maths and I do the calculation of the mathematic experience here, if I take the upper end of the inspection, so 10 and 10 percent of the time, it’s going to need replacement, which means it would cost me 100 to inspect in order to find a defect, which needs to be replaced. Replacing cost me, if not, I take the lower end of , that range 250 per meter. So even if I’m in the absolute worst conditions and all the factors down, it still has a 2. 5 X factor, which makes it worth it to inspect every single time before I take any rehabilitation decision. So that sounds like pure evidence.

Is it that straightforward? Like you would always inspect and assess before you go to works or would sometimes people just say, you know what? I know my network and I can listen to it. And I know I need to dig here.

Amanda Siqueira: No, they definitely put a camera into the pipe. and so there’s oftentimes a delay between that first inspection that identifies a defect that. does need a rehabilitation of up to six months. And so the contractor that gets awarded the rehabilitation will sometimes say, well, hang on, this footage is six months old.

I don’t know what I’m going to get myself into here.

I might lose a camera down, so I’m going toput a camera in again, then they’ll do the rehabilitation itself, and then they’ll put the camera in again to show evidence to the water authority, yep, I’ve done this and so therefore you can pay me now.

So there’s lots of footage getting captured.

Antoine Walter: So drum roll, we’re 45 minutes in and we need to answer that question.

Business Model and Revenue Streams

Antoine Walter: Where do you make your money?

Amanda Siqueira: it’s mainly from the decision support. So being able to say to people that, you know, we go to councils and municipalities and they say, look, I’ve only got 200, 000 this financial year. But they’ve got, a 2 million backlog. And so they’re what do I do?

How do I spend the money? Best. And so we’re able to reverse engineer a lot of the asset health information about their network that they’ve already inspected, or sometimes that they’ve inspected now and say, well, In priority order, this is what are the highest risk assets. And if you tell us you care about cracks, then all of them become high risk.

But if you tell us you actually care about fractures and maybe only the breaks or infiltration then we can start to whittle it down to a more manageable number. That means that you can go to your council, get sign off, you can get it out to contract really quickly. And start to whittle down that infrastructure backlog instead of accelerating it further by spending the money in the wrong

way.

Antoine Walter: I get the value you’re bringing, but how do you collect your fair share of that value? What’s your business model?

Amanda Siqueira: , we work on a per meter or per foot basis. So every meter that’s processed or every foot that’s processed of inspection footage that goes through our platform we have a rate associated with that. So it’s basically to say that at a certain tier level.

So let’s for argument’s sake, a thousand or 10, 000 meters there’s a certain rate associated. And if you go above, then the rate drops. So the more that you use the kind of cheaper it gets. And so we, you know, bias for volume and make sure that we’re aligning value with delivery.

Antoine Walter: so you’re solely. Paid on the amount of footage, which you assess if they save billions. Thanks to you. You don’t take a cut.

Amanda Siqueira: Correct. We’re looking at pricing. We’re looking, well, I mean, this is the challenge with the value metric and it’s, it is genuinely it’s not just us, but I think it’s not, pricing is never easy thing to to nail down. But our current value metric is per foot or per meter of inspection footage processed.

We’re looking at, you know, different ways of identifying that because, you know, . Some organizations increase or decrease their inspection footage year on year. , and we know that they know when we know that we’re delivering value. So, how do we look at that differently?

So it’s a constant food for thought.

Challenges and Market Expansion

Antoine Walter: So now I need to take my hat of being the bad guy. What would be the number one objection? Like You go in and they tell you not interested because of that. What’s that one? And then I have a second one.

Amanda Siqueira: actually something that you brought up earlier, which is, I don’t want the. Like the granularity from my network, like I’m happy with what I’ve got. I’m happy with the level of data that I’ve got. I don’t need to be able, I don’t struggle with meeting budget and the asset health information that I’m getting from my current provider.

So, in that case, thenthe value proposition degrades.

Antoine Walter: I think that’s the saddest time I was right. so that means that they would not go for it just because they prefer being blind and playing it like the ostrich and having the head in the ground.

Amanda Siqueira: . In markets where the asset management policy is quite strong and there’s been consistent over time prudent spending of on high risk assets and their backlog is small and they’ve got a good proactive program they don’t need as much of a solution like VAPAR yet, you know, we’re expanding our offering, but our ICP, our ideal customer profile is very much those organizations that are woefully underfunded and really require a data solution to help create a pathway out to reducing that infrastructure backlog for them.

Antoine Walter: So that means public utilities, because those would be slightly more underfunded with the exception, of course, of the UK kind of weird animal, which is a bit special. so your ideal customer profile would be those public utilities, but you would not go to I guess Denmark or Singapore where they are rightfully funded.

So why would they? Need to optimize

Amanda Siqueira: most of our work comes out of the UK because there’s a very weird scenario of privatization in that industry. But the fact remains the level of funding does not meet the requirement of the network. So, that’s true very much in the UK because of the combined sewer system.

Exists also very strongly in the U. S. So we’re seeing a lot of exciting things come out of the U. S. Where we’ve got, say, in Australia, I’ve got a separated system. So, the sewer pipes and the stormwater pipes are completely different. They’re not the same pipe. There’s not as stark a difference between how much funding people have and the infrastructure backlogs.

Antoine Walter: you didn’t deliver to Paris, for instance, because we are recording the third day of the Olympics and who saw the opening ceremony knows that there was a heavy rain on Paris, which means that now all the open water. Sports are in trouble because yeah as many billions that they poured into the network, I guess they haven’t inspected the right pipes at the right place and it is a combined sewer and now it’s not meeting quality, but would those be also. Like, there’s a big event L. A. is coming up in four years, maybe they want to be better prepared than Paris. Would that be a use case for you where there’s high stakes or really it’s about infrastructure management?

Amanda Siqueira: So there’s a number of drivers, such as the Olympics. even when the Olympics were in Sydneya lot of driversEnhance the amount of investment that goes into the pipe network. Sydney built a 14 kilometer tunnel storage situation interceptor that basically stopped any of the sewage go out into Sydney Harbor for the games.

So, you know, probably something similar that Paris is invested in there is a very cool, if you, I don’t know if you’ve been to it, there’s a very cool sewer museum right next to the Eiffel Tower. You should definitely hit it up, I’m sure Paris has done something similar, LA as well.

Antoine Walter: Those drivers are also very helpful, but the big tailwind that is, You cannot ignore is the the depreciation of our assets the requirement, the infrastructure backlogs, just globally going up. That’s more of a of a slow burn driver for us that is definitely going to help us with adoption. Totally off topic. I’m sorry about that. Do you still have in Sydney that four kilometer pipe, which goes into the ocean because it was faster to build than upgrading the waste treatment plant, which they did for the Olympics, because that way you could have all the sports going on the coast

 I wouldn’t be surprised if it was still there. there’s so many solutionsthat need to be employed last minute to take care of these problems rather, but yeah, whether they’re longterm is the

Amanda Siqueira: question.

Antoine Walter: Okay. Sorry for the sidetrack. Bring you back on track. You mentioned how the UK is one of your main markets. Is it a challenge as an Australian company based out of Sydney? I know that you have also an office in the UK, but time zones. are against you is that troublesome or for an AI startup? Come on.

Everything can be remote.

Amanda Siqueira: unfortunately we haven’t figured out how to make AI align time zones. So yes, it is a very big challenge. In any organization, people are the biggest asset in a, in an organization. And so it does create a unique Kind of scenario. And I think people will have to be very good at communication.

Cause like I say, nine hours is the best case actually in winter. It’s 11 between the East coast of Australia and and the UK, , we just got our first team member in us. So it’s 24 hours, which is fun. And again, just like elongates those communication timelines between the team.

You mentioned the U S on your website, there’s this thing which says us one reference more to come. So is it like your next elder radio? After all, it’s much larger than the UK and arguably they have At least the same kind of trouble, but how challenging is it for you? I mean, what I’m trying to understand here is that you just raised 500, 000 from investors led by Pureterra Ventures.

Antoine Walter: And I’m just out of my little brain thinking that after all, that’s not so much money. So if you want to take on different geographies, the entire world, can you do that with that amount of money, or do you need to focus and say, You know what the U. S. is it, and when I say the U. S. it’s those four states on the east coast and let’s start there and let’s build our beachhead.

Amanda Siqueira: the latter part of what you just said is pretty much the strategy is to make sure that we’re able to cement ourselvesto get ourselves into the U. S. And then make sure that we’re got, you know, at least 5 to 10 really happy customers to start with get ourselves to a milestone as well.

So generally when startups fundraise, you’re raising to a milestone. You’re raising generally 24 months is a rule of thumb of runway. And there’s, many schools of thought on this, whether you raise as much as you can or raise as little as you need.

And we’re raising for a milestone raising for that , two year runway to make sure that we’re in the U S we’ve got really good customers. We’ve got good traction, but also that , we’re also in the UK. That’s our biggest market. that regulatory period that, you know, it was just about to begin.

We’ve got to make sure that we’ve cemented our position in the UK it’s very exciting. There’s a lot of innovation happening there. So, yeah the five mil is going to be, it’s going to be pushed.

Antoine Walter: So your philosophy is to raise as little as you need.

Amanda Siqueira: Well, we raise as much as we need. Cause I think that’s what any startup would be doing. Just to make sure that you don’t want to over dilute, but you also want to make sure that you’re not stifling the company. Just because you want to protect your position.

Antoine Walter: In your UK story, you said you were facing that other AI company. And when I was doing my research, I found out that Nowadays, every specialist water venture fund has its own J well in the crown in terms of AI assessment of sewer and stormwater lines. And so for future adventures, it’s going to be you.

But I think Burntisland Ventures is investing in sewer AI, and I can’t remember who’s the one investing in fluid robotics, , but I’m just wondering being several different companies in that field, I would see that as a positive because that means there’s a market and it’s much easier with a lot of brackets around the easier build the market.

If not alone on the other end, Sure, AI is a US company backed by US investors. So they’re, if you go to the US, you’re going to fight the locals and it can be tricky.

Amanda Siqueira: Yeah, definitely. it’s very encouraging for us to see the traction that you sir, AI having fluid robotics as well. it gives us a lot of really positive market signals in the US and tells us a little bit of a story about the success that we might enjoy.

And that’s what we’re bidding for, right? And so, yeah, the challenge, there’s definitely a challenge. But yeah, we’re hoping to rise

to it.

Antoine Walter: Do you talk together like there would be a big villain, which is traditional approaches and you would be all fighting that big villain together. Or is it like pure competition or you don’t even that much even notice yourself because the water industry is so scattered that you can really be living in parallel timelines without ever crossing.

Amanda Siqueira: , the U S is So big, right? There’s so many water utilities and municipalities in one state alone, you could have, thousands and there’s like little pockets of people that know each other and things like that. So, Yeah it’s a very big place. There’s lots to be done there.

And each organization is coming at it from a completely different angle, albeit, you know, very similar technology. So we’re all using AI. But. You know, as an example, fluid robotics has a hardware component to what they do. So AI has a they don’t do the hardware themselves, but, you know, they interact with hardware certainly more than we do.

And for that reason, their customer and their ICP is slightly different to us. The problem that they’re solving is slightly different. Albeit the technology is very similar, ICP is can be slightly different in some cases.

Antoine Walter: So coming back to your own fundraise, I mean, there are several ventures payers and Pure Terra is the lead investor, but there’s one which is not a venture, which is Autodesk. So you have Autodesk on your cap table. What does that tell us?

Amanda Siqueira: Yeah, we’ve, we are absolutely stoked to have Autodesk on our cap table. Autodesk over the last couple years are heavily investing in water and we’re absolutely thrilled. So they canvas the market in the AI kind of pipe space . And selected VAPAR from the vendor that they wanted to proceed with in terms of just partnering.

And then, you know, it started there and then they, we had this race coming up and so they kind of put their hand up and said that they were interested to learn more. So, we’re super excited to have Autodesk. They’re doing a lot of investment in water and for audit for us?

you know, a global brand the customer footprint that they have in North America.

You know, we’re bidding for market share forhuge amounts of distribution, help and sales assistance that we can enjoy now because we have this partnership with Autodesk

Antoine Walter: Autodesk somewhat created the first ever water unicorn. acquired Innovize for 1 billion.

So. My thinking process is when they agreed on the valorization, Innovice became a unicorn. And then when Autodesk acquired Innovice, then they were not a unicorn anymore because they were acquired. , but just to say they are ready to, go big on, or they seem to be ready to go big on water topics.

So is that what you aim to be building? Or is it like totally unrealistic to say that a CCTV AI inspection tool can become a unicorn.

Amanda Siqueira: So if we were to stay in in exactly what we’re doing now I think that would be a very big challenge to make that claim. But yeah, no, our ambition is is. Much much higher than just doing pipes. So yeah, very excited to own own this pipe space.

So what are you building in the next five or 10 years,

Future Vision and Product Development

Amanda Siqueira: our whole kind of company and product vision is that it’s actually about public assets. So When we think about the customer in the ICP, the ideal customer that we’re trying to serve, it is the asset owner. So people that are trying to maintain these public assets with taxpayers money that just, You know, chasing the tails because the investment is potentially going to the wrong areas, or they don’t have enough.

And so the problem that we solve now and speaking about is related to simple pipes and drainage pipes. But there’s a lot more asset classes than just those two that this problem relates to. in the next couple of years, we’ll be looking at translating what our workflow is and the models from more than just sewer pipes more than just drainage pipes to adjacent ancillary assets and how we can grow that from there.

So, it is going to be, you know, that’s, that in and of itself is very crowded space. So we have no illusions that’s going to be an easy one, but such is the challenge of the role.

Antoine Walter: that’s, the vision and I’m, as I’m already going far too long and I’m sorry about that, I won’t use that open door you’re giving me but I’m going to mark that for the future because that’s a red ocean. That was Platformization and being like the OS of the water utility. I think that would make for a great sequel conversation.

So I put that one in the fridge, but I still have one question, which is that’s your vision. Now, if you translate it into your action plan for the next six or 12 months. What needs to happen?

Amanda Siqueira: Yeah. So what we’re doing now for water utilities is saying what’s gone wrong in your pipe. We’re using AI to deep learning over the inspection footage directly to say, here’s a lot of effects here and here. And then we use another bit of software to say, okay, this is what based on what you’re telling us, this is what you need to do about it.

We’re also helping people understand when they need to do something about it. So overlaying contextual information, whether it’s geospatial or risk and then cost. So when do I need to do something? And how much is it going to cost me?

Those two. moment I handled kind of cause I on platform. And so what we’re trying to do is complete that workflow to say, you’re able to come to a decision completely within our product on these four fundamental questions that help you fix an asset before it fails or fix pipe before it fails.

And once we’re able to do that in a very seamless way Then, you know, we’ll look at expanding out to to other asset classes.

 I thought that would be the last question in the deep dive, but now I need to ask, how do you get your insights on how your customers behave with the tool? Do you proactively go with them and look at what they do? Yeah, Well, so there’s this theory for like product management, but just even products called jobs to be done. And it’s about looking at. I think the, I don’t know if there’s a equivalent hardware store in France, but in Australia, it’s called Bunnings. And it’s basically the analogy of somebody walks into this hardware store.

And even though they ask for a quarter inch drill well, they actually want is a hole in the wall. But then hang on a second, they don’t actually want a hole in the wall. They want a painting there. if you were a Bunnings or a hardware company I could sell them a drill.

I could sell them a hook like a, you know, the three M ones that so there’s a bunch of solutions that could solve the problem. If you understand the core reason as to what the job is that they’re doing. And so. We look at our customers and foster a relationship with our customers such that we really try and understand their jobs as well, or if not better than them, so that we make sure we’ve only got one productone platform.

And when we release something, it has to work for everyone. So we want to make sure that we understand the core reason why people are doing what they’re doing and align it with everyone so that you know, can’t make everyone happy. Butthat deep understanding of the customer drives how we build.

Antoine Walter: Well, again, that is another area where I would probably push further, but I think that’s a sign. We need to have a sequel. Thanks to Amanda for everything you shared in that deep dive. To round off those interviews, I have a set of rapid fire questions. If that’s fine for you, I would switch to that. And in the meantime, about that jobs to be done story, there is an awesome episode of the everyone hates marketers podcasts on that very specific topic of jobs to be done.

I’m going to link it in the show notes. And for now, the rapid fire questions start with

It’s time for the Rapid Fire Questions!

---

Rapid fire questions:

Antoine Walter: what is the toughest challenge in your opinion for water tech startups?

Amanda Siqueira: Sales cycles.

Antoine Walter: What would be your best single piece of advice for the founders and managers of the about 1000 early stage water startups?

Amanda Siqueira: Understand sales cycles.

Antoine Walter: What’s the drop of knowledge you wish more investors knew about the water sector?

Amanda Siqueira: I think how sticky contracts can be and how once you build that trust with a customer genuinely build that trust and relationship, there’s a huge amount of long term value.

Antoine Walter: So if I take a sad track from the two first answers, I understand that the cost of acquisition is super high because of the long sales cycle and all of that, but the lifetime value is also super high. So it’s worth the effort. Do I get you right?

Amanda Siqueira: Correct. But a lot of investors just say, Oh, six months sales cycle or like 12 months sales cycle. and they’re just like, see, yeah,

Antoine Walter: What was your most unexpected partnership and what did it bring you?

Amanda Siqueira: our most unexpected partnership was actually with a competitor who I won’t name, but It was very early days. We explored and actually got quite far down the line of co developing something together. It taught us a lot about that organization, but also about partnering what are some of the gates that we should put in place for no go or go, no go for partnerships and things like that. And so yeah, it actually really helped us partner in the future, which is an unexpected consequence.

Antoine Walter: You got me curious, but you said you can’t name them. So I respect that super short profitability or growth.

Amanda Siqueira: Gross.

Antoine Walter: What is the next profile you’ll hire

Amanda Siqueira: Us sales

director,

Antoine Walter: when you hire, are you looking for sector experience or startup experience?

 Sector over startup. Like specialist over startup. Opening new markets or doubling down on the current ones? And I think your previous answer tells us a lot about that.

Amanda Siqueira: Yeah. But but yeah, opening, opening new markets I think is. the focus of the two, though both are very important

for us.

Antoine Walter: What is that tool nobody speaks about, but you couldn’t live without?

Amanda Siqueira: Google tasks,

Antoine Walter: Oh I got Google drive, Google calendar in that section, but never Google task. So Google has a whole suit of things that people couldn’t live without. Interesting.

Amanda Siqueira: I mean, spreadsheets really, like, I feel like spreadsheets just in general. That’s also I’ve got like very cool spreadsheets that I love.

Makes you a little bit less of a cool kid just because you answered that. But yeah, sorry.

 oh no,

Antoine Walter: What’s the single piece of insight your ideal customer profile needs to hear right now?

Amanda Siqueira: They probably know the reality that not only you know, as it’s getting worse, but the other insight is we just don’t have enough people in the industry to do this sort of work. We can’t train people to sit there and watch CCTV for eight hours a day. And the people that are doing it now?

To be blunt.

Um, And so it’s, there’s this, silver wave of people that are going to start retiring and take that really fundamental and important knowledge hopefully not with them, but may do so, It’s an inevitability. You know, we’ve doubled our doubled our numbers on many accounts, but certainly usage year on year I’m sure others in the space that can say to similar numbers you know, people are really starting to look into this there is a better way.

And so definitely find out more if you think that you need some help.

Antoine Walter: What are you desperately needing and want to raise an open call for right now?

Amanda Siqueira: So very excited about getting our product into the U. I think there’s a huge need for it. We’ve seen the uptake in the UK which is a similar market. So yeah very keen to hear from any municipalities or water authorities in the North American area.

If you want to learn more about what we’re doing please reach out at vapidocco, CEO.

Antoine Walter: As always, those links will be in the show notes. Now, caveat as well, you know, the UK has two handfuls of water utilities. The US has 155, 000 of which 50, 000 must be wastewater utilities. So yeah, it’s a it’s a scattered iceberg you’ll have to take on. Last question. What can and should I do for you?

Amanda Siqueira: What can you do? Oh, you’ve already done so much, Antoine, even just having me on this podcast like I said, I’ve been a massive fangirl of your podcasts and your content online. So, definitely keen to stay a follower, but yeah, if there’s anything else that I can help with or.

keep on delivering your content.

that is straightforward. I’m telling, I’m asking you if I can help you and you’re telling me keep doing what you’re doing and I can help you. That’s a good deal. If all my conversations were that easy, you know,

but generally I think there’s not enough of a conversation about. water technology, startups the funding the issues in the space. And so really your content at least strikes according with me, and I’m sure that’s the case with all of your other followers. So, definitely you keen for you to keep doing what you’re doing.

Antoine Walter: I really appreciate I mean, it’s been a pleasure to spend that bit more than expected time with you. If people want to follow up with you, what’s the best place for me to redirect them to?

 Yeah. So they can check us out on our websitewww dot vapa co. Or hit us up on LinkedIn as well. Off track which is absolutely stupid, but one hour, you know, I can allow myself, I picked dww. show for my website. And then a lot of people are asking me, is it wrong? Because it should be www and you say dww. So it must be wrong. And you pick co, which people would be thinking, Oh, you missed the M, right?

Amanda Siqueira: Yeah.

Antoine Walter: But yeah, it’s on

Amanda Siqueira: Oh man. It’s the worst. Someone’s domain sitting on vapid. com and I’m not bothered to pay them. So I’ll just sit there with CEO.

Antoine Walter: So as always, the links are in the show notes, so check them out. Amanda, it’s been a great pleasure. I stand my point. I would be super happy to have a follow up with you. The next time you hit a big milestone, for instance, when you get your traction going in the U S or whenever you feel like it’s worth it to share the story, because we have still open doors from today.

Amanda Siqueira: So I can imagine that if we feed that with more in the future, it’s going to make for another cool conversation. So thanks a lot for that and talk to you soon. Thanks, Antoine.

Other Episodes: