- next.js
- convex
- clerk
- typescript
- tailwindcss
- locationiq
- vercel
Project background
I co-founded Global Pet Sitter with my friend Jo. Two developers with over 10 years of housesitting experience between us, trying to build something transparent and community-driven in the pet sitting space.
The problem
After years of using platforms like TrustedHousesitters and Nomador, we got frustrated with the direction these platforms were heading:
- Fees getting out of control — booking fees on top of annual memberships felt like nickel-and-diming the community
- Customer support that doesn't respond — emails go unanswered, just chatbots when something goes wrong
- Review systems that feel broken — false reviews can't be removed even with proof
- No real vetting — platforms take minimal action even when theft or pet neglect is reported
Something that started as community-focused became over-commercialised. We thought we could do better.
The tech stack
We went with Next.js 16 (App Router + Turbopack), Convex for the real-time backend, and Clerk for authentication. Tailwind for styling because life's too short to write CSS from scratch.
Why Convex? Real-time updates out of the box. When a sitter applies or a message comes in, you see it instantly. No websocket setup, no polling, no headaches. The schema is defined in TypeScript, queries are type-safe, and deployments just work.
Building trust
When you're asking someone to take care of their pet, you're asking for a lot of trust. Pets are family. We spent way more time on verification, reviews, and safety features than initially planned.
Blind reviews
Both parties submit their review independently, and neither can see the other's review until both have submitted (or 14 days pass). This prevents retaliation reviews and keeps things honest.
Review importing with AI
Experienced sitters already have reviews on other platforms. They don't want to start from zero. So we built a review import system:
- User links their profile on another platform (TrustedHousesitters, Nomador, etc.)
- We verify ownership by having them add a unique code to their bio
- Once verified, they upload screenshots of reviews
- GPT-4 Vision extracts the review data automatically
Features
- Map search — find sits in your area visually with marker clustering
- Dual accounts — be a sitter and an owner with the same profile
- Privacy controls — you decide what info to share
- Real-time messaging — no refresh needed, messages just appear
- Blind reviews — honest feedback without fear of retaliation
- Review importing — bring your reputation from other platforms
- Early adopter rewards — founding members get special lifetime pricing
Lessons learned
The tech wasn't the hard part. The best code in the world doesn't matter if nobody uses it. Two-sided marketplaces are hard — you need pet owners AND pet sitters, but each side won't join without the other.
Every complaint is a feature request in disguise. Every confused user is showing you where your product falls short. Listening changes everything.
The impact
- Building a growing community of pet lovers who believe in transparent, community-driven pet care
- Members importing reviews from established platforms, proving the value of portable reputation
- Real-time features creating a smooth, modern user experience
- iPhone app currently in development
