It all begins with my mom asking me reviewing her skincare product. I didnt found genuine reviews , she was relying on tiktok reels which might be biased and she was completely lost. Identical ingredients, different brands, different prices. She looked at me like I had answers, I didn’t. I have no idea what skincare was. I was speechless. But that confusion stuck with me.

I started paying attention. The skincare world was massive, way bigger than I had realized. People were obsessed, scrolling through Reddit threads, instagram/tiktok reels at midnight trying to figure out which serum wouldn’t destroy their skin. There was clearly a gap. And yes there is still out there. So I did what any curious builder would do. I sent out a survey to friends, asking about their skincare struggles, what they wished existed, what frustrated them.

The survey changed everything

The results caught me off guard. Almost everyone I knew was into skincare. Even the friends I never expected. That was the moment it clicked. I needed to build something, a platform where people could get genuine product reviews, generate personalized skincare routines, join forums, and access real resources. I was convinced it would be a game-changer.

Building solo, nights and weekends

I was still working full-time and still today. Most of the building happened on weekends, unless I had on-call duty. Solo. No co-founder, no team. Just me, taking a leap and seeing where it would lead.

The ideas kept growing. Why not generate skincare routines based on a user’s specific context? Maybe even real-time face scanning or photo uploads to personalize recommendations further. I went deep. I even visited a dermatologist myself. They handed me a prescription and I had to buy a bag full of skincare products. During the checkup, I asked them about data, whether we could use anonymized patient information to build something meaningful. I might be in the wrong place to ask it. In the US, there are compliances to navigate. In Nepal, I had never heard of anything like that. No one was recording or classifying that kind of image data. They had billing records but nothing I could use.

Public skincare datasets did exist on platforms like Kaggle, but validating them for real-world use and training a reliable model was a challenge on its own. It comes with cost too. Then came the harder questions, how do you handle compliance and user privacy when you’re capturing real-time face data or accepting photo uploads? Every edge case carried real risk. I had to think through all of it, the security, the liability, the potential warnings, before writing a single line of code.

Gallery

The lessons that hit hardest

Here’s what I learned the hard way: don’t try to do everything at once. Focus on one thing. Iterate. And above all, talk to your users. Talk to them early, talk to them often, talk to them right now. Them means right audience.

When you build shiny features in isolation, without real user involvement, you end up with a product built for yourself, not for the people who actually need it. Before you write a single line of code, have absolute clarity on why you’re building, who you’re building for, and what problem you’re actually solving.

The skill I never knew I was missing

As a technical person, I used to think the code was the hard part. It’s not. The most underrated skill for any builder is are marketing, sales, communicating to real users, and iterating based on what they tell you.

I didn’t have those skills when I started Skinfem. I’m still learning them. It’s overwhelming. But the things you pick up along the way teach you more than you’d expect. Way more.

And I want to be honest, sometimes building yourself up is a deeply personal choice, not just about earning or career gains. It’s about becoming someone who can handle more, see clearer, and build better next time.

In the end, I stepped away from Skinfem to sharpen these skills in my current role, where I can apply them at a different scale. A win is a win. And this chapter, even though it’s closing, made me a better builder than I was before.

I have to improve my communication and starts speaks upon camera.

As of now, the AI routine generation feature has been completely shut down. I can no longer fund it. The remaining functionality will likely go offline by the end of February not all of them, and the domain probably won’t be renewed. Skinfem will fade out quietly, but what I built and learned from it won’t.

On giving up

I don’t see this as giving up. Failures are just things that happen along the way. They’re not the destination. I’ve failed at things before, things I cared about deeply, and every single time I came out the other side knowing more than when I started. Skinfem is no different. What matters is I showed up, I built, I learned, and I moved forward.

Giving up would be stopping entirely. I’m not stopping. I’m redirecting…

If you’re building something similar in the skincare or health-tech space or anything, I’d love to hear about it. I’m happy to share deeper insights from my experience, brainstorm ideas, or even explore collaboration. Feel free to reach out via an email. Always happy to help a fellow builder.