How do you build a social platform for pet owners from scratch — fast enough to validate the idea before running out of runway?

01 — Problem

Pets City was an early stage startup with a warm idea and a cold clock: a platform connecting pet owners with verified breeders and service providers, a tiny founding team, no existing infrastructure and a limited runway to find out whether anyone wanted it. I joined as the full stack engineer, which at that size means the technology choices, the architecture, the implementation and the deployment pipeline were all mine to get right or wrong. The trap in every greenfield build is spending the runway on foundations for a product that might pivot next month. The opposite trap is hacking something together that collapses the moment real users arrive. I had to walk between the two.

02 — Constraints

• Startup velocity — architecture had to enable fast iteration without accumulating technical debt that would slow us down once traction arrived
• SEO was non-negotiable for discoverability of pet listings and services — server-side rendering wasn't optional
• Single engineer making calls that would be expensive to reverse: framework, data model, auth strategy

03 — Architecture

Next.js was chosen from the start for SSR — pet listings and breeder profiles needed to be indexable by search engines on day one, not after a JavaScript-dependent crawl. TypeScript across the full stack caught contract mismatches between API responses and frontend state early, before they could silently corrupt user data. The styling layer used SCSS with a lightweight custom design system rather than a component library — less surface area to fight, faster to tune per feature. The whole thing shipped as a PWA so mobile users got near-native install behaviour without a separate app.

04 — Solution

I built the platform from an empty repository: Next.js with server side rendering so pet listings were indexable by Google from day one, TypeScript across the stack so API and UI could not silently drift apart, a lightweight SCSS design system instead of a heavy component library, and a PostgreSQL data model for pets, listings and users. I shipped it as a PWA so mobile users got an installable, app like experience without anyone funding an App Store build. I wired Mailchimp for early signup capture and Vercel for continuous deployment, then ruthlessly prioritised what could be validated fast: pet profiles, listing creation, basic connections between owners and breeders. Everything that needed serious infrastructure waited for proof that users wanted it. Working directly with the founders, I turned an idea into a live product people could install and use.

• Full-stack Engineer — greenfield product build
• Next.js SSR architecture, TypeScript codebase, SCSS design system
• PWA setup, Vercel deployment pipeline, Mailchimp integration

05 — Outcome