The starting point
The HoReCa sector (hotels, restaurants, cafés) in Moldova has constant staffing needs, but the large, generic job boards represent this niche poorly: listings for waiters, cooks, or baristas get buried among thousands of other industries. On top of that, the labor market is trilingual — Romanian, Russian, and increasingly English — which makes generic job boards even less effective here.
HorecaJob.md was built to close exactly that gap: a focused platform, exclusively for the HoReCa industry, in all three relevant languages.
The approach
The project was built solo — from data modeling to deployment. Key decisions:
Trilingual from day one, not bolted on later. Instead of building one language and translating afterward, the routing and content system was designed for RO/RU/EN from the start. That avoids the typical rework that retrofitted internationalization usually requires.
Over 40 SEO landing pages, generated programmatically. Instead of a single generic "jobs" page, landing pages were built specifically for concrete search queries — combining job category (waiter, cook, barista) with location. That's the lever that makes organic search traffic possible at all for a new, unknown platform.
A full application flow, not just listings. Employers can post jobs, candidates can apply directly through the platform — no detour through external emails or forms.
Technical stack
Next.js, TypeScript, Tailwind CSS, Prisma, and PostgreSQL. The combination of server-side rendering (important for SEO on the landing pages) and a relational database for listings, applications, and user accounts was the obvious choice here — no overengineering, no microservices for a project of this size.
Result
144 commits, built solo in three weeks from the first line of code to launch. The platform runs in production at horecajob.md.
Why it matters
None of the decisions above are arbitrary — each maps to a cost or a growth lever a real business feels directly:
Trilingual from day one means the platform never needs a costly retrofit to open up the Russian- or English-speaking side of the market later. That's addressable market, decided in the architecture instead of a future sprint.
40+ programmatic SEO pages are the acquisition channel. A new platform with no ad budget and no existing audience has essentially one way to get its first users for free: rank for the searches they're already making. That's what makes a project like this viable without marketing spend.
Applications handled on-platform, not via email, keeps both sides of the marketplace — employers and candidates — inside a funnel you can measure and improve, instead of leaking into inboxes you can't.
This is a self-initiated project, not a paid engagement, so I don't have public traffic numbers to share yet. But the reasoning above is exactly what I'd apply to a client's platform — and it's testable: organic traffic, application completion rate, and employer retention are the numbers I'd track from week one.
Have a similar niche idea, or need a multilingual platform? Let's talk.