Bartek Pucek and Zbigniew Sobiecki are enabling companies to build proof-of-concept apps and integrations for their prospects in minutes.

Robin: Hello, both of you. Thank you so much for joining us today. Bartek, could you start by telling us a bit about your journey up to now?

Bartek Pucek: I'm Bartek, CEO and co-founder of Proofs. I've been building online products for many years and started very young. As a teenager, I began building things in various capacities, from startups to scale-ups and even in enterprise environments. I'm also an active angel investor in companies like Eleven Labs, Kick, and ZetaLabs, among others, both in Europe and the US. Additionally, I'm a newsletter author, sharing my learnings about being a founder and investor; my newsletter is now among the top 20 most popular Substack newsletters in the business category.

What were you doing right before you started Proofs?

Bartek Pucek: Before Proofs, Zbigniew and I were at Forward Operators, an AI lab where we developed solutions for various companies. These included notable names in different sectors, such as Ten Square Games, Google, and Decathlon.

Zbigniew, could you share your background and how you ended up in this project with Bartek?

Zbigniew Sobiecki: Yes, I'm Zbigniew, co-founder and CTO of Proofs. I've been building software since the late '90s and am still hands-on. However, I grew tired of traditional software and was intrigued by AI. With my experience as a CTO and co-founder, integrating technology and product development, Bartek and I considered what the next big thing we could build with this technology would be, and that's how we got here.

Bartek Pucek: That's an understatement because Zbigniew is very modest. Before Forward Operators, he was pivotal in building Quick-commerce, a leader in Poland, and he embodies what you'd call a '10x engineer.' But more importantly, if you measure a person by how much they help others grow, ask any engineer who has worked with him—they'd confirm it.

Zbigniew Sobiecki: Thank you!

Zbigniew, do you have anything you want to add about Bartek?

Zbigniew Sobiecki: Bartek is fantastic and complements my skills well. It's beneficial to have someone who can move fast but also delve deep when necessary. We've been working together for quite a while, and it's been a great partnership.

Cool, and how did you both get involved in AI initially?

Bartek Pucek: It depends on how deep you go. If you consider AI pre-ChatGPT in terms of data science and machine learning, it's been part of software development for many of us. But, after our introduction to ChatGPT, we decided to establish an AI Lab to learn more.

And how did you come up with the concept for Proofs, and how does it work?

Bartek Pucek: This didn't start with one of us waking up one day with a brilliant idea. It began through our work in the AI lab, talking with different companies and founders. We kept hearing about difficulties in building on top of APIs and showcasing their value. This problem piqued our curiosity, and for us, it grew into an opportunity to try and build a product. We aim to change how software is built and sold, integrating AI agents into the process.

Can you elaborate on how Proofs works?

Zbigniew Sobiecki: Essentially, we're building end-to-end projects for companies from a single prompt, particularly in the sales process. When you have a significant lead, the best way to convert them is often to build an app specifically for them, showing how their commerce or enterprise can integrate with our stack. However, this is time-consuming and doesn't always lead to a conversion. So put simply, Proofs automates this process, allowing projects that would take hours, days, or weeks to be showcased in minutes, enhancing creativity and improving customer conversion.

Bartek, why this product and why now?

Bartek Pucek: If you see an opportunity to build something that fundamentally changes how software is delivered to customers, there's no turning back—you have to pursue it. If you can visualize that end goal, there is no other way once you recognize it. You can't unsee it, and you just need to build it. This doesn't guarantee success, but it compels you to try. Regarding why now, I believe it's still early for AI agents and code generation, but things are beginning to align in terms of building a more effective process for delivering and selling software compared to the inefficiencies that exist today.

Moreover, there's a lot of talk about AI co-pilots increasing personal productivity. For example, if I'm using an AI co-pilot, I can be 20% more productive. However, if five people are using co-pilots, it doesn't mean that the company's productivity increases by 100%. For AI agents; if you can manage an end-to-end process efficiently with an AI agent, what happens is transformative. In our case, you will be able to deliver products and integrations to your prospects not in months or weeks, but in minutes, and not at the cost of hundreds of thousands of dollars but significantly less, with a remarkable impact on sales margins and tailored more closely to their needs than ever before. Ultimately, this reduces all inefficiencies in the process. There's a lot that needs to align today, not to deliver on that vision immediately, but to build towards and possibly achieve that vision in the coming years. We are building for what we believe will become the standard go-to-market approach for software in the coming decade.

So, can you tell us what your plan is for the next five years?

Bartek Pucek: Honestly, I only really know what we’re building in the next two or three weeks; I don't have a clear picture for five years. However, we believe that when you want to buy customized software solutions, based on code and APIs, you won't go through the current process of multiple meetings and a long sales cycle. It's inefficient, costly, and time-consuming just to try something, not to mention building your product on top of someone else's solution. We envision a future where you ask an AI agent to build it for you, and it gets done very quickly. We hope for Proofs to be the backbone of that process for companies building and selling that software. So, if there's a tangible vision for the coming years, it is to build that part. And go far beyond a PoC.

Can you talk a bit about what the role of the developer will be with an AI agent? How would developers in companies work with your product?

Zbigniew Sobiecki: That’s a great question and one we get asked quite often. We're building something that eliminates the minutiae and frustration of repeatedly building the same things that don't advance your career. Similar to how LLMs might impact journalism—taking over tasks like writing standard comparison reviews that people might not want to write—our product does the same with proofs of concepts. We are making these processes relevant and showcasing the product without necessarily building it line by line. The role of the developer is going to change in that you won’t need to focus on the minutiae but can instead concentrate on the bigger picture. The technology is already here, and while it won’t eliminate developers, it will empower them significantly.

Thank you, Zbigniew. Bartek, I wanted to talk briefly about your investing and your role as an angel investor over the past couple of years. What does that look like and what do you look for when you’re investing?

Bartek Pucek: Okay, so the most successful investments I've made involved a blend of urgency and the capability to deeply dive into the topics they're tackling. They strive to understand the problem from a first principles perspective, consistently pushing out products and soaking up non-stop feedback from potential customers—literally around the clock, 24/7. From an investment standpoint, I always ponder whether I would back the same team if their initial venture failed. If they go bankrupt with their first idea, would I invest again? And if that second idea failed and they pivoted to a third, would I still be on board? Honestly, the best investments were where my answer was an unequivocal yes—I'd invest regardless of initial setbacks because these teams and founders were relentlessly resourceful. There's a mantra in angel investing: kiss your money goodbye, yet be ready to double down again and again if you believe in the team's ultimate success. It doesn't matter if it’s their first or fifth venture. Eventually, a stellar team will pinpoint a significant problem and excel in solving it, making any investment worthwhile.

Please tell me what the future looks like for Proofs, your team, and whether you're hiring more people in the next couple of months. What are your big goals in terms of building your company?

Zbigniew Sobiecki: Yeah, so I think the future looks almost blindingly bright sometimes, which I'm sometimes wary of. We've built a great team; we have around eight people, and all of our engineers have 15 to 20 years of experience—they're all software veterans. Regarding the future, echoing Bartek’s earlier point about focusing on the next two weeks—things in AI are progressing so rapidly it’s impractical to predict the long term. There are versions of the future on the horizon that we feel will be great, but who knows what will happen in the next few months. Our immediate goal is to build as many sophisticated and complex projects as reliably as possible, and that’s what we’re striving for with Proofs.

What gives you both the courage to go down this route without really having a playbook? What drives you?

Zbigniew Sobiecki: What drives us is the excitement of the unknown and the challenge of it—tackling what seems almost impossible but just possible enough to attempt. That’s what motivates me, at least. Otherwise, what’s the point?

Last question. Zbigniew, what are the things we need to be thinking about for the future in terms of making sure that this development is something positive but also brings about new things in a good way?

Zbigniew Sobiecki: Absolutely. I think there's going to have to be a human level that's very important with AI. If I'm worried about anything, it's that on one hand, people try to solve everything with AI without really thinking about what the problem is, and then we risk losing some of the abilities we've had before. So, there are significant social implications to consider.

Great, thank you both so much for your insights!