Mobile App · UX/UI
d.i.t
d.i.t. is a two-sided marketplace that connects people with vetted experts (ditters) for live 1:1 video sessions: a bridal makeup trial, a wardrobe refresh, a cooking walkthrough. Two connected apps, one AI assistant keeping scope, prep, and bookings clean. Designed end to end and wired into an interactive prototype.
Problem
Expertise is either free and generic or expensive and inaccessible.
Someone who wants to look right for a wedding or finally cook a dish properly has two bad options: hours of one-size-fits-all video content, or booking a professional across DMs, scheduling links, and separate payments with no trust signals. Most intent never converts.
Experts have the mirror problem. Stylists, makeup artists, and coaches struggle to monetize depth; social platforms reward reach. The 1:1 tools that exist are built for business advice, not visual, personal, appointment-like services. Five frictions kept surfacing: noisy discovery, clumsy booking, mismatched scope, thin trust, and leaky monetization.
Buyers with a specific need
Want their exact question answered live, not a generic tutorial: a trial, a fitting, a walkthrough.
Experts
Stylists, makeup artists, cooks, and coaches with deep personal expertise and no clean way to charge for live time.
Repeat relationships
Both sides once a session lands: save, rebook, buy the products the expert actually used.
Ideation
Product decisions built around trust between strangers.
The design was shaped by a series of decisions, each resolving a tension between simplicity, fairness, and trust. Pricing became a hybrid menu: session types at 15/30/45 minutes with a base fee plus per-minute overage, simple to set up but fair when sessions run long. Booking landed on a hold: payment authorizes immediately, the ditter confirms within 24 hours, and the card is only captured on confirmation.
Star ratings punish subjective work, so public profiles show curated testimonials, sessions completed, and a would-rebook rate, while the star score stays private for matching. The AI assistant was scoped as a helper, not a gatekeeper: it flags scope mismatches and coordinates prep inside a three-way thread, but never decides for anyone.
Solution
Two connected apps with one AI assistant.
d.i.t. turns an expert's time into a bookable live product. The buyer app runs four tabs (Discover, History, Saved, Profile), with the revenue path flowing from task intake through expert selection, AI-assisted scoping, payment, prep, the live session, and outcome. The ditter app mirrors it (Home, Sessions, Earnings, Profile) with onboarding, availability and calendar sync, the session-type menu, and hold/confirm requests.
Describe
The buyer states the task; guided intake and the AI classify it and route to the right ditters.
Book
A session type is chosen, the AI checks scope against length, payment authorizes, and the ditter confirms the hold.
Go live
Prep notes arrive in the three-way thread, the session runs on video, and the outcome flows into testimonials, rebooking, and product links.
Features
The hold
Bookings enter a tentative hold the ditter confirms within 24 hours. Payment authorizes now, captures on confirm; nobody gets ambushed.
- Structured decline reasons feed matching
- Full refund if declined
Testimonials over stars
Public profiles show curated testimonials and objective signals: sessions completed, would-rebook rate. Stars stay private and power matching.
- Fair to subjective, visual work
- Past sessions are opt-in, never auto-published
An AI that helps, never gates
A three-way thread with the buyer, the ditter, and the assistant. It flags scope mismatches, shares prep notes, and keeps coordination out of DMs.
- Scope validation before money moves
- Prep checklists per session type
Earnings that read like a statement
Ditters see upcoming sessions, requests, and a bank-statement-style earnings breakdown with payouts, all in Roboto Mono.
- Menu pricing set once, edited anytime
- Two-way calendar sync, opt-in
Screens
Viability
Focused on personal services, not business advice.
d.i.t. sits at the intersection of the creator economy (50M+ people monetizing knowledge) and online personal-service verticals where live guidance is going virtual. The demand signal is sharp: 46% of virtual beauty consultation takers purchase within three hours, which is why product links ride along with every session.
Existing 1:1 platforms (Clarity.fm, Superpeer, Topmate) prove people pay for live access but treat the call as the whole product and skew to business advice. d.i.t. wraps the session in AI scoping, prep, and a trust system built for subjective work, and monetizes at 30% per session plus an attribution-friendly product layer.
Outcome
Both journeys fully designed and prototyped end to end.
Both journeys are fully designed and wired end to end: ditter onboarding through earnings, buyer discovery through outcome, the AI thread on both sides, and the cancellation/refund framework. Two launchable prototype flows cover the complete demand and supply experiences in the beige/olive system, in both color modes.
- Trust design is product design: the hold, private stars, and opt-in privacy did more for the concept than any screen.
- Specifying every screen's states and navigation before Figma made a two-app system buildable by one person.
- Next: Finalize refund percentages, extend AI prep automation, and explore group sessions.
Stackt