Hackathon limit: 3:00. The video is recorded on the live site (https://strandlab-youcam.vercel.app) on desktop (1920×1080) and a phone viewport (390×844, shown in a device frame). The voice-over is a young Indian English male voice (Sarvam AI Bulbul v3, speaker "rehan"). The background music is original, synthesised in code, and contains no copyrighted material. The YouCam APIs are named out loud and listed on the closing card.
This file is generated from videos/script/timeline.json. Rebuild instructions are in the Demo video section of the root README.
| Time | On screen | Voice-over |
|---|---|---|
| 0:00–0:15 | Problem: the guessing loop (animated) | Walk into any hair-care aisle, and the labels expect you to already know your hair. Is it 3B? Is it fine? Is it high porosity? Honestly, most of us just guess. So we buy a product, it doesn't work, and we buy another one. |
| 0:15–0:22 | StrandLab home page | Meet StrandLab. It measures your hair with YouCam's Hair AI, and turns those numbers into exactly what to buy. |
| 0:22–0:39 | Scan: guided camera, 4 photos | You take four photos: front, both sides, and one with your head tipped forward. The guided camera tracks your face live, tells you out loud how to move, and clicks the photo on its own. Every photo is auto-framed for YouCam's detectors, so you never get stuck. |
| 0:39–0:57 | Live analysis progress, porosity quiz, dew point | Four YouCam APIs run in parallel: Hair Type, Hair Frizziness, Hair Density and Hair Length Detection. Meanwhile, I answer what a camera can't see, like porosity. And StrandLab checks the dew point in Mumbai, because humidity changes which ingredients actually work. |
| 0:57–1:17 | Results overview: measurements, takeaways, weekly routine | Here's my result: 3B to 3C curls, high density, and a humid week ahead. So: rich products, hold for my curls, and skip glycerin on humid days. Every rule shows the measurement behind it. And nothing is a black box. You can see every YouCam call, and every photo that auto-framing rescued. |
| 1:17–1:32 | Ingredients tab, Routine tab | The Ingredients tab is my label cheat sheet: what to look for, what to avoid, and why, based on my measurements. The Routine tab turns it into four wash-day steps: cleanse, condition, leave-in and style, each with its reason. |
| 1:32–1:40 | Coach answering in Hindi, voice in/out | Confused by any of it? The coach explains it in plain words, in 17 languages. You can even talk to it, and have the answers read out loud. |
| 1:40–1:52 | Products: ranked kit, real-product search, label photo verdict | Then, shopping. A ranked starter kit, and real product links for your own country. Standing in a store? Just photograph the bottle's ingredient list, and StrandLab tells you if it suits your hair. |
| 1:52–1:59 | YouCam Hair Volume try-on: realistic vs ad promise | And YouCam's Hair Volume try-on shows what this routine can realistically do, right next to the ad promise. |
| 1:59–2:18 | Demo store: widget, match badges, events | Now, the big one for e-commerce. Any online store can add StrandLab with one line of code. The shopper taps "Find my hair match", matches once, and every product on the page shows how well it suits them. The store gets the result as an event. And the shopper's photos never leave StrandLab. |
| 2:18–2:32 | Phone: install PWA, home screen, camera, offline | StrandLab is also a PWA. Install it from your browser, and it opens full screen, just like a real app. It's on your home screen in the store aisle, uses your phone's camera, and still opens when the network drops. |
| 2:32–2:46 | Phone: no-photo check, text size, high contrast, mic, share | No camera, or don't want to share a selfie? The no-photo check builds your plan from four picture questions. There's bigger text, a high-contrast mode, voice in and out, and you can send your plan to your stylist instantly. |
| 2:46–2:57 | Close card: pillars, URL, YouCam APIs used | Your hair, measured by YouCam, explained in plain words, and ready for any store. This is StrandLab. Try it free at strandlab-youcam.vercel.app |