April 10, 2026
Magnetic AI
MagneticAI's goal is to give men real, data-driven feedback on how attractive their photos are, by predicting the opinions of thousands of women (filterable by age, ethnicity, interests, personality, and more).
Portfolio | Melt Studio

Melt Studio built the React web version of Magnetic AI's live iOS app in four weeks, as part of its pro-bono prototype program — giving founder Jesse Jaffe a web product to show investors before his fundraise.
Magnetic AI: an AI photo-rating app for men's dating profiles
Jesse Jaffe founded Magnetic AI to show men what women would think of their photos. Women are paid to rate photos of men, and an AI model trained on those ratings — around 20,000 of them — gives each photo a score from 1 to 10, with specific things to change. The iOS app was already live and in the market.
The challenge: a web app before the fundraise, not after
Jesse was raising money, and investors kept giving the same answer: not enough traction, not enough revenue, and only on iOS. The plan had been to build a web version after the round. But the round needed traction, and traction needed web. Jesse isn't a technical founder, and his small team was busy keeping the iOS app running.
The solution: turning a live iOS app into a React web app
Jesse didn't need a rebuild. He needed a second way into the product, and he needed it fast. So we scoped it that way:
We built the web version in React, using the live iOS app as the reference.
We left the iOS app and the existing backend untouched.
We shipped the flows that matter: sign-up and login, photo upload and AI scoring, galleries, a paywall, and Stripe subscriptions.
We left Android and a broader platform migration for later, when the business calls for them.
Along the way, our team found and patched a security vulnerability that had gone unnoticed.
How a dedicated team shipped it in four weeks
A dedicated team with a project manager and our CTO, weekly demos, and a shared Slack channel. Jesse set the priorities as Product Owner. We handled the engineering.
Results: 12 features, 37 user stories, and a new B2B revenue path
12 features and 37 user stories delivered in four weeks, approved by Jesse as Product Owner.
A working web app before the fundraise, not after it.
A new revenue path: personalized models for matchmakers, a B2B product that runs on web.
"Having it before the fundraising is really a game changer for us." — Jesse Jaffe, Founder, Magnetic AI
Have a live product and not enough engineering bandwidth for the next step? That's the gap we fill.
Schedule a call today!
Thx for reading. With Love, MS.




