SÉRA

I have combination skin. I figured that out by googling symptoms at midnight. Before that I spent years following skincare influencers who all said something different, buying products based on Amazon reviews written for someone else's skin type, and wondering why nothing was working.

The problem was not a shortage of information. There was too much of it. Every brand had a routine to sell. Every influencer had a stack to recommend. Every Reddit thread had fifty conflicting opinions. None of it was personalized. None of it accounted for how my products interacted with each other. None of it gave me a clear answer.

I started asking friends and realized this was not just me. People were spending real money on routines that were not working and had no idea why.

My dad was the one who pointed out that people do not just want to know if a product is safe. They want a full routine. Morning and night. Something they can actually follow.

That unlocked the second half of SÉRA.

The other moment that stuck with me was from my own experience trying to find the best SPF. I had six tabs open. Every site had a different product at the top. Every review contradicted the last one. I closed all of them and just bought whatever was on sale. That is not a research problem. That is a decision problem. SÉRA is the answer to that.

SÉRA has two core flows, both built around the same goal: give people a clear answer fast.

Check your routine. You enter your existing products and SÉRA analyzes each ingredient, flags conflicts and harmful chemicals, and gives your routine a body burden score out of 100. The score is grounded in EU AI Act standards and the EWG Skin Deep database. If two of your products are working against each other, SÉRA tells you exactly which one to drop and what to replace it with.

Build your routine.You answer four questions about your skin type, primary concerns, budget, and climate. SÉRA generates a complete morning and evening routine with safety scores and current prices for each product. The recommendations aren't based on sponsored content or affiliate relationships. They're based on ingredient compatibility and safety data.

The hardest design challenge was making the scoring feel trustworthy rather than arbitrary. Anyone can display a number. The hard part is making the user believe it. Every decision in SÉRA pushes toward explainability: you can always see why a product was flagged, what specific ingredient caused the conflict, and what standard it was measured against. The verdict comes with a reason. Without that, it's just a number on a screen.

I started SÉRA at SheBuilds 2026, the Lovable x Anthropic x Stripe hackathon in March. The brief was open. I had four hours and a problem I'd been annoyed by for years. The first version was rough but the core was there: enter your products, get a score, understand why.

I shipped it at the end of the hackathon and moved on. Then I kept thinking about it. A few days later I called my dad on video chat and asked him to go through his actual skincare products with me in real time. He grabbed everything from his shelf and we went through them one by one. When SÉRA flagged a conflict between two products he'd been using for months, and suggested a swap that cost about a third of the price, something clicked. He said, "this is actually useful." That was the moment I knew it was working.

I went home and kept building. Added the routine builder flow. Rewrote the explainability layer so the reasoning was always visible. Made it faster. Fixed the edge cases that showed up when real people used it. Added support for more product categories. Improved the ingredient database coverage. Each version was tested on real people who had no obligation to be kind about what wasn't working.

Launched it properly at seracare.app. Early users from UW, friends, and family have tested it, and the feedback has been overwhelmingly positive. People feel like it actually understands their skin. That's the goal.