The Rise of Lean Product Teams: How We Built an Entire Product as Just Two Engineers

I think due to the rise of AI and the adoption of vibe coding, development teams are becoming increasingly lean, shaped more by experience and domain expertise than by headcount. I’ve been working on a product of my own with just one other person. We are a two-man team: I handle everything on the backend, along with DevOps and networking, while my partner focuses entirely on the frontend. We designed the product, defined the features, laid out the architecture, and broke the deliverables into phases, all on our own.

Between the two of us, we cover the full spectrum of product development, from idea to execution. We're not just engineers, we’re also capable of handling business analysis, UI/UX design, and product ownership. Both of us have prior experience in owning and delivering products end-to-end, which allows us to think beyond just code. We approach every decision, whether it’s a user flow, API design, or infrastructure strategy, with the mindset of building something usable, scalable, and valuable. With our combined skill sets, we can take a product from the drawing board to a production-ready, scalable solution, without needing any additional resources.

Now, when we bring AI into the mix, things get even more interesting. While AI certainly accelerates development, that’s not the primary value it adds. More importantly, AI enables a new kind of collaboration, a way to brainstorm, evaluate ideas, and get quick feedback on implementation approaches. The code it generates may not always be production-ready, but it often serves as a strong foundation to iterate upon. By maintaining a healthy balance, leveraging AI as a creative and technical assistant, rather than a code generator, we’re able to move fast without compromising quality.

Our setup is a good example of how small, highly skilled teams can deliver end-to-end products. Because we cover all essential roles, engineering, design, analysis, and ownership, we don't need a large team. As long as each core function is handled by someone with deep expertise, a lean team can build and maintain highly functional, production-grade applications.

This model has broader implications. Organizations will likely take note and begin optimizing their teams for efficiency. The drive to reduce overhead and increase velocity may lead to a shift where fewer people are responsible for more. This could pose challenges, which I’ve discussed in a separate blog, but it’s the direction things are heading.

What we’re also starting to see, and will likely see more of, are compact teams of friends or ex-colleagues forming around shared ideas, building real products without needing to scale the team unless necessary. These teams don’t wait for permission or funding. They build. However, in cases where such teams lack expertise in areas like scalability or infrastructure, their projects may stall. Many of these products may end up living in GitHub repositories, which are technically sound but lack the push to production-grade scale. In our case, we are planning to push it to GitHub and open-source it because our goal was different. We were experimenting, and our goal was to see if we could build a product end-to-end, but if we get a good response, we certainly have the capability to go build a complete and scalable application, but that might not be the case for everyone

That’s where organizations come in. These lean, well-designed prototypes can be picked up and scaled by companies with the resources and operational muscle to take them further. In a way, we may start seeing a new ecosystem emerge: small teams as idea incubators, with organizations acting as accelerators for the most promising outcomes.