Cofactory

Context

Cofactory is a unique company founded in 2023 by two Stanford AI researchers. The name stands for "Company Factory". It aims to create thousands of companies in the next few years, leveraging AI both as a tool to build products much faster and as an engine to power the products themselves.

Contribution

I joined Cofactory as a full-stack engineer a few months after its inception. I worked on one of their first products - an AI writing assistant for structured documents - which is set to become the company's first foundation product for its AI product explorations. My role has been to develop this product from scratch and prepare it for user tests.

I have worked closely with the founders and was primarily responsible for the product's frontend direction, implementation, and some backend features, namely the Stripe and GPT-4 API integration. Having never played with OpenAI's API before, Vercel AI SDK did a great job at abstracting the initial complexity, making it easy to build a custom API to serve the generated content to the frontend.

Challenges

Cofactory is a young company with a very ambitious vision, currently in the exciting phase of finding its product-market fit. As a startup eager to validate its business hypotheses and raise money, shipping regularly and iterating fast is critical. I believe we've done well in this aspect: Kostantin and I worked hard and managed to roll out our first product version in just two weeks, following it up with new, improved versions every fortnight.

However, this pace of execution came with its set of issues. I've noticed a tendency to equate speed with success in some early-stage startups. Speed without clear direction can create an illusion of progress in the short term, only to create more significant problems down the line. My approach has been to keep the codebase consistent and maintainable and strive to find clarity in the product direction. At times, it's felt like swimming against the current.

I've sometimes struggled with the scope of my position. Hired as a full-stack engineer, I often found myself dabbling in product management and mentoring junior developers, as my role was shifting week by week. I enjoy these tasks, but the constant change made building momentum in any of them difficult. Reflecting on it, I should have proactively voiced this concern with the leadership team earlier.

The (in)famous startup mantra of moving fast and breaking things might have worked for some over the last decade. In today's environment, where competition is fierce and users expect a polished experience, building a solid foundation for your product is more important than ever. I believe in the slow is smooth, smooth is fast motto, focusing on consistency and momentum over sheer speed and quick fixes.

Results

I believe I left the product in a good state at the end of my contract: a clean codebase with clear conventions and a solid foundation for future development. But I understood that some companies are not ready to invest in these things at such an early stage, validating the product-market fit should remain the top priority. I'm looking forward to seeing how Cofactory takes this first product to market and what other companies they build on top of this one.