The Code Stitcher | Anand Kulkarni from Crowdbotics

DISCLAIMER: Below is an AI generated transcript. There could be a few typos but it should be at least 90% accurate. Watch video or listen to the podcast for the full experience!

Anand Kulkarni 0:00
Want to make it really easy for people who are facing the problems that we can solve for them to find us and that’s how we’re approaching the system.

Alexander Ferguson 0:15
Welcome, everyone to UpTech Report. This is our Founders Journey series. UpTech Report is sponsored by TeraLeap. Learn how to leverage the power of video at Teraleap.io. I’m excited again to be joined with our guests. Anand Kulkarni, there we go, who based in Berkeley, California, CEO of Crowdbotics, definitely check out part one of our interview where they, we dived into the actual platform of, of being able to build apps without coding, using this machine learning machine intelligence to combine 1000s of packages from cross open source universe to create ready to use products. But this discussion, I want to dive a little bit deeper into your story. I mean, how did you get to where you are today?

Anand Kulkarni 1:00
Yeah, well, I’m happy to give you the the big the big beats of my career. So originally, I was a academic researcher, was a PhD, researcher at UC Berkeley had an NSF fellowship in mathematics, did a bit of academic work published a lot of papers, I spun out an enterprise software company, out of grad school with three other grad students as my business partners, and ran that for six years, we were working in the human computation, machine learning and sales technology space. So that was my first exposure to running that is called Lead genius. Yeah, we were, we were venture backed, and, you know, sold product to a lot of big companies. It was a it was a, it was a fun, a fun company to build. And, you know, we had a phenomenal team over there. Around four years ago, I became really excited by what was happening in machine learning. And, you know, when I was in grad school, you know, I did a lot of work in optimization and machine learning, but it was regarded as a kind of a dead end, stagnant field at the time, hard to believe now. But, you know, I was in a wall and people said, yes, it’s, you know, it’s, it’s hard to do better than what we’ve got right now. And, you know, we started to see really impressive results from recurrent neural networks and LSM networks, and essentially, the entire deep learning movement, and I was really impressed by the idea that we could create things, I was seeing academic results that were creating things that looked like code, using machine learning systems. They weren’t exactly code, but they looked like code. And I became really, you know, impressed and said, Okay, we got to I got to work on this. This is the this is going to be the way that people build in the future. So we set out to build a company that could do that. Of course, you need a lot of data, lots of modules. As I mentioned, modules are the way that software engineers like to build already by finding packages, components and tools that they can snap together to get things off the ground quickly. And so that seemed like a natural place to start. And of course, over the course of those years, we found there was lots of demand for people to build software, which which is a permanent tailwind, I think in our current era. So that’s been the the journey to get us to where we are today.

Alexander Ferguson 3:31
Now, getting getting getting started one at one thing is to have the idea. The other thing is to be able to have the funding to be able to make it happen. I’m curious, on the funding side, what have you seen as some common state mistakes, or even a biggest mistake one could make when seeking funding?

Anand Kulkarni 3:47
Yeah. So I think there’s two ways that people can raise raise money, right, you can raise money based on your justification concept and your resume. Or you can raise money based on traction. And out of the two, I have found that it’s always easier to raise money based on traction, meaning having Do you have customers? Do you have users? Do you have technology built and it’s massively easier to raise funding with that second one, especially if you have revenue than it is to raise money without it. And of course, you can go either way. Lots of founders, you know, raise pre revenue. But in my experience, both advising companies and building companies myself, I think it’s always a sound approach to look at the fundamentals of your business and say, Hey, we’re we are already generating cash flow. There’s money coming in. We know how to make this work. We just need to take it bigger and that’s an easier conversation to have. Then then saying believe us.

Alexander Ferguson 4:50
Yeah, then please leave us for you guys. Have you gotten funding yet? Have you in bootstrapping? What’s your

Anand Kulkarni 4:55
Yeah, so we so we bootstrapped in the beginning. Up to Just under our first million dollars of revenue. And at that point we raised, raised our seed round from some really phenomenal people from Hunter walk at homebrew and Michael Dearing and Harrison metal and the folks over at be partners and UC Berkeley’s house fund, as well as some really smart angels. So yeah, we have raised money. And I think that we were lucky to be able to work with some really talented people who joined the joined the company at that time,

Alexander Ferguson 5:39
or building the team and being able to to grow from there. You kind of have to have the right people, right, like to be able to keep growing people are so crucial. curious of your experience, what are some common mistakes you can make when it comes to hiring?

Anand Kulkarni 5:56
Yeah, okay. So I remember 12 years ago, when I was first looking at hiring people for the first time, and I said, How do I know that somebody is really good, you know, I should just hire people who are my friends. And, you know, say, let’s, let’s make it work. But it turns out, you can be systematic about hiring. And by developing a really well honed sense of how people perform, you end up making really outsize and having really outsized impacts in your company. So some of the things that we look for here, of course, we want to make sure people are technically extremely sound, we have a very technical product. And it’s important that people who are building that product can understand all elements of that. So typically, that means a track record, historically, of having worked on very diversity of interesting technical and difficult things, often in a architectural or leadership capacity. And then familiarity across multiple frameworks and multiple languages. Generally, all the software engineers here are what we call polyglots. They speak speaking multiple languages. Meaning, you know, we’re not we’re not a pure JavaScript platform or a pure Python platform. And so we have to make difficult decisions about the trade offs in different languages and development models, because our customers inherit those choices from us, right, when you spin up an app on crowd biotics, you are going to be getting a lot of code that was generated by the platform. And it’s important that that code is up to snuff. So when we hire our team, we make sure people are fluent in that, in that approach, that they can move from one system to another without really losing their, their confidence. It’s

Alexander Ferguson 7:46
having that polygon is the ability to understand the multiple languages and the technical aspects. But working well as a team, imagine as you grow, the culture fit is also important, and how you manage that growth is crucial. How big is your team today?

Anand Kulkarni 8:02
So we’re 30 people full time. And yeah, the culture the culture is is crucial as part of it can’t it can’t call excellence, but you know, it’s it’s also in there.

Alexander Ferguson 8:16
Are you guys located in different cities? What’s Yeah,

Anand Kulkarni 8:20
I mean, I think since since COVID, our practice has become the norm. Yeah. So we are distributed across the US and in part across the globe. But you know, I’m here in Berkeley. And my head of engineering is in LA, my head of product solutions is also in LA, we’ve got a couple folks in Austin, somebody in Nashville, someone in New York City, a few folks in Mexico City, so it goes outward from there, I think that the great thing about being able to be remote friendly, especially in a post COVID world is that we can hire folks from absolutely anywhere in the world. And that means you can tap into the best talent, regardless of if they happen to be in your neighborhood or not.

Alexander Ferguson 9:08
Having the right team allows for growth. And it sounds like you guys are definitely experiencing the the big need and demand for for this type of solution. I found it interesting actually started off describing you were focused on really developers and like, just making it easier for developers to be able to code faster by just already having these packets ready to be able to pull in, but then you realize there’s a bigger, bigger need. I’m curious how how did that change or morph your thinking when it comes to marketing and being able to get the word the story itself? Any lessons learned there?

Anand Kulkarni 9:44
Yeah. Okay. So one thing that’s worth mentioning is that, you know, most of our most of our business, if not all of our business is inbound, meaning people come and discover us by reading our materials on the web, finding our blog content, finding technical content that we’ve put out there. which is great. That’s a phenomenal way for us to grow. But you know, that shift is really driven by the fact that we realize that business users are the ones who have big needs around this, right? Developers are great. And obviously, whatever we build has to have strong buy in from technical users, because they always sign off on decisions that are being made inside organizations. But, you know, developers who end up framing their problems a little bit differently, when they go into things often like to tinker, experiment, try new technologies. Often there’s a long period between when they first encounter attack, and when they say, let me actually use this for a project. And I think that’s the the reason that we shifted, when we ended up building a business, we saw a lot of demand coming in from people who were not developers and said, Hey, there’s there’s an opportunity here, which looks like the right move.

Alexander Ferguson 10:56
How do you plan on? Or do you? Have you seen marketing and today’s environment where it’s very distracted, everyone’s doing digital marketing, because you can’t do in person events or conferences or trade shows? How have you seen a good way to get attention, especially to business leaders? So those that are the key decision makers that are needing your app? How do you break through the noise?

Anand Kulkarni 11:18
Yeah, it’s a good question. So um, you know, a lot of our strategy is built around the idea that when you’re building an application, you can look at that space of low code is kind of a, you know, a black box on the outside, there’s lots of solutions and lots of providers. And you don’t really care about that, as a business leader, all you actually care about is a solution to your specific problem, right? So I need to build a dashboard that does this right now, let me find what dashboarding solutions are out there are that are HIPAA compliant, right? That’s a good example of where you might start your search process. So for us, we have thought about looking at a series of niche problems. And looking at that universe of niches as the place that we direct our marketing efforts. And there’s some ways that our product is uniquely well suited for that because we have lots and lots of applications built so far and a lot of different domains. So you know, there’s this history and track record of kind of usage in different areas. But I think that’s how we think about it. You don’t necessarily want to at least not for us, we’re not trying to make a concerted effort to blanket the marketplace, we want to make it really easy for people who are facing the problems that we can solve for them to find us. And that’s how we’re approaching the system.

Alexander Ferguson 12:39
How do you go ahead about measuring the customer experience? Yeah, so someone comes on?

Anand Kulkarni 12:46
Yeah, great question. So I think the methodology that we apply is one that should be standard and look familiar to a lot of low touch SAS companies. So we have an inbound funnel, we score everybody who shows up in that funnel, depending on what they’re doing. Basically, how successful were they trying to build an application? How much did they engage with various articles, that we put out technical articles around how to build, and then what parts of the product that they using, then, of course, we’ve actually got places throughout the product, where we make it really easy to go and ask for help. Right? So if you’re trying to get something custom built, you can’t figure it out, okay, there’s the button to click to talk to a person and we’ll, we’ll connect you into that pathway. And you know, that, that approach, making it kind of easy to get help is one of the ways that we have thought about helping customers scale up more quickly and get more success out of the platform. But yeah, you know, the approach of inbound user scoring and kind of keeping track of who is getting stuck and who is not, you know, I think that’s the approach that user friendly SAS platforms really ought to take, when they are trying to understand you know, how to make their user experience better.

Alexander Ferguson 14:05
Or expanding and you got the initial interest and things are moving, how do you know when it’s, it’s the right time for expansion and whether it’s bringing in funding and to truly actively go after customers versus just letting inbound content come? What’s your indication and methodology there?

Anand Kulkarni 14:23
Okay, so, so I will clarify that inbound is inbound is controllable, if you’re driving marketing appropriately, right. So, you know, we’re watching specifically how much content we’re producing every month and also how much traffic we are deploying through performance marketing into those content locations and into our core properties. So there is it’s not to say that inbound means that your growth is organic. It means that, you know, we have carefully tailored our strategy and our marketing spending towards making sure inbound works really well for us. Now the question is, where do you know the time is right to try and grow? So if you’re a funded startup, a venture back startup, you are always asking this question of how do we grow even faster than we are growing right now by deploying our capital more intelligently? So that’s the decision you make, ideally, before you raise your first dollars, but definitely every time you go out to get more capital, you’re asking, you know, can we grow faster than we’re going? Now? Do we understand our economics well enough for us, because our marketing is very predictable. We know that spending a certain amount of dollars on user acquisition results in a certain number of customers showing up as paid users at the end of the funnel. And for any founder who’s looking at this question, you should be looking at those metrics from day one, as we have for you know, how much is a customer worth? On average? How much is a user worth? On average? Who’s free? How much can we spend to get a user and still have it come out profitable for us? And then how quickly can we scale that up without breaking the channel? And that’s the sort of the art of building an internet business in in 2020. It’s really more of a science,

Alexander Ferguson 16:15
the alchemy of Okay, let’s see where this where this fits. How long did it take you to find those measurements that that metrics?

Anand Kulkarni 16:22
Yeah, so we so the first the first part of the business that we built up, you know, before we raise money we had, we got to initial traction, just based on let’s call them informal mechanisms, right? So pounding the pavement and hustling over the course of the next year after that, when we took capital, we continued onward, and started collecting data religiously about our cost of customer acquisition, or user value, our conversion metrics and our conversion funnel, and now that’s a key stat for us, we track that every day and look at those numbers on a daily basis to understand where are people getting stuck. And the nice thing is, this corresponds directly to where people are having difficulty in the product, right? So if people are stuck in creating an app, they can’t figure out how to get their goal achieved. Yeah, they’re never going to turn into a customer, right? They may not even succeed as a free user. So that tells us that’s the place we need to go. And we need to look. And that also tells us where we should be deploying capital, right? Should we be investing more in customer acquisition to pull more users in? Or should we be spending that money on engineering instead to go fix the place that the product is giving people friction? And there’s always a balancing act between you know, are we good enough to justify more acquisition now? Or should we invest more in product improvements that lead to a better product experience for our users? And usually, you want to do both, right? Pulling customers make your product better? And you know, it’s a growing cycle.

Alexander Ferguson 18:02
Looking forward from here, what challenges do you see you’re going to need to overcome in order to keep scaling and growing?

Anand Kulkarni 18:10
Yeah, okay. So we’ve, we’ve scaled a pretty to a pretty good level this year. And I think we’ll hit pretty comfortably. Our targets for 2021. You know, even even though this was a difficult year for most of the market, we didn’t actually experience that a lot of folks ended up needing software and more as a result of virtual Yeah. And so we’re looking at a great 2021 ahead of us, which is fantastic. I think that the the impediments that we’re looking at here and the things that we think about, first, can we continue to provide enough coverage on the types of solutions people are looking to construct, that we maintain the advantage that we have right now, in providing good solutions for customers who are searching for problems that they’re trying to solve? The second one is, can we make sure that our end users find it easy enough to create an app that we’re able to convert those those four users into successful app creators early on in the process? Because that’s the basic metric that our customers look to us to achieve. And our users? Look, you know, they’re coming to us to figure out how to build. So we think about like this time to first app metric, right? How fast before you signed up after you’ve signed up for the system? Are you seeing a running app in your, you know, in your web browser or your interface? So we want that to be under five minutes? Because, you know, that shows that we’re doing our jobs, right. So, you know, bringing that number down to be as small as possible is, is really our key key product metric to look at for user success. But yeah, 2021 I think looks looks like a great year to scale up. You know, we’re seeing a lot of market demand for what we’re doing. And of course, a performance marketing strategy based on inbound customer acquisition scales up really nicely. In a lot of ways,

Alexander Ferguson 20:14
as CEO and leader what books, audio books, podcasts, have you listened to our reading right now that you would recommend?

Anand Kulkarni 20:23
Yeah, okay. So there’s a bunch. Good one is, if you’re a first time founder, high output management by Andy Grove, who was the CEO of Intel for a long time, and basically breaks down management as an engineering problem. And it actually explains that software development can be made predictable, predictable, which is anathema to a lot of technologists. But you know, he argues that yeah, actually, you can break it down into steps and decompose it and it works out really well. So that one’s great. Jason Lemkin runs blog called Sastre, which is, you know, now now a juggernaut in its own right. But if you’re building a SaaS company, it’s a lot of educational content there. And then, of course, I am a massive fan of the material that’s put out by Y Combinator, the tech accelerator. I’ve participated in YC, myself. And I will say that, I think they’re they’re reducing the practice of building startups, technology startups into a science. So it’s really a worthwhile set of materials that they put out from documentation and legal docs, advice on how to process and structure around even thinking about your business all great, great content to tap into.

Alexander Ferguson 21:43
Last question for you, what kind of tech innovations do you predict we’ll see in the near term, the next year or two and long term, 510 years from now?

Anand Kulkarni 21:52
Okay, so language models for artificial intelligence, GPT, three, and the related category of things that GPT three represent are a massive disruption and a massive disrupting force that the market has not yet recognized and appreciated yet enough. But you’re going to see over the course of the next five years, increasing numbers of places applications where we can harness language models to talk to interfaces, or talk to systems in ways that we have not, we would not have believed possible in the past, right. So I’m extremely bullish on what natural language models are going to do. Obviously, we’re doing a lot of stuff, thinking about how it’s going to impact impact software development, but you can imagine it’s going to impact lots of domains where you end up interacting with human beings today, their jobs are going to become much more efficient, because they’ll be able to use those, those tools to help. I think there’s also the big potential for using those tools in creative contexts, right? Art, Media, writing production, all places where GPT three is going to be able to kind of change the role of a content creator in some basic ways, right? Instead of somebody sitting sitting down and writing articles entirely by hand, maybe you’re going to write the outline, pass it through GPT, three, or a language model a few times, and you know, end up using that as kind of a predictive tool to amplify your output, or better explore the space of possible creations you might make. And so there’s going to be a massive impact with this. I think that it’s been under hyped, to be honest, compared to compared to where it is. And that’s the that’s the nature of AI. Right, as you know, people always debate is it going to be that impactful or not? And then by the time that it actually gets into applications, people say, Well, okay, it’s not really that impressive. It’s just, you know, Google, Google, Google Home, you know, giving us perfect audio responses, you know, so these are the kinds of innovations I think we’re gonna see. There’s going to be big, big changes in how we use technology products that will surprise us.

Alexander Ferguson 24:07
I really appreciate the your passion and excitement for the future and that it’s not hype. If anything, it’s, there’s more to this. What’s possible with with voice enabled AI, I like it. So thank you so much. And now for sharing your journey and the lessons learned here. There’s a lot of great insight.

Anand Kulkarni 24:28
Hey, excited to talk about this. I think it’s great stuff.

Alexander Ferguson 24:32
Now definitely for those that want to learn more about Anand’s adventure Crowdbotics, go to crowdbotics.com and you can sign up for a free trial check more out. Also see part one, our first part of the interview where he shared a deep dive into what they are doing. Today again, our sponsor for this episode TeraLeap. If your company wants to learn how to better leverage the power of video to drive sales and marketing, head over to TeraLeap.io and learn about the new product customer stories. Thank you so much, guys, and we’ll see you next time. That concludes the audio version of this episode. To see the original and more visit our UpTech Report YouTube channel. If you know a tech company we should interview you can nominate them at UpTechreport.com. Or if you just prefer to listen, make sure you subscribe to this series on Apple podcasts, Spotify or your favorite podcasting app.