Lungu Andrei

Hey StartupBase 👋 I built Folicle because tracking hair-loss treatment with your camera roll is basically useless — every photo is a different angle and light, so you can never tell if minoxidil or finasteride is actually working. Folicle aligns each new scalp photo to the previous one, follows a Hair Score over time, and exports a clean, dermatologist-ready PDF for your next appointment. It's intentionally non-diagnostic — it documents progress, it doesn't diagnose or prescribe. iOS + Android, waitlist open. Would genuinely love your feedback.

The local/Ollama mode is the detail that makes this genuinely usable for sensitive files — a lot of people can't push client docs or photos through a cloud API, so an offline one-time-license option removes the main blocker. Custom naming instructions are the other smart call: bulk renaming only helps if it follows my convention (date-client-version) instead of the model's guess. Does it also read EXIF/created-date for photos, or purely the visual content?

Auto-generating the zoom effects from actual interactions is the part that saves the most time — manually keyframing zooms in post is exactly the tedious step that stops people from polishing their demos. Can you tweak or delete an individual auto-zoom afterwards if it lands somewhere you didn't want? The aspect-ratio and cursor controls are a nice touch too for turning one recording into both a landscape and a vertical cut.

The "decision engine" framing is what stands out to me — most analytics tools surface the data but still leave the actual spend call to a human. I'd be curious how Cortex handles cross-channel attribution before it recommends reallocating budget, and whether it acts on that automatically or just flags it for the team. Optimizing directly for CAC is a smart wedge too, since that's the number that actually decides whether early acquisition is sustainable.