META5Software engineering studio
currently buildingbackend platforms & APIs
META5 builds the backend platforms, web and mobile products, IoT fleets and cloud infrastructure that businesses across Europe and the US run on. Nine years of it, from a first commit to the pager that never goes off.
/capabilities
We take the whole path — architecture, build, deploy, and the on-call rotation after. No hand-off to a team that has never seen the code.
Services that stay predictable when traffic is not. Schema-first data models, hard boundaries, and refactors that pay for themselves.
Refactored an enterprise platform down to 50 ms responses.
Interfaces people use every day, on the web and in their pocket, built on one shared type-safe core instead of three diverging ones.
Architected Austria's 5th-largest e-commerce app.
Hardware in the street, talking to software you can reason about. Device messaging, provisioning, telemetry and remote recovery.
40 services moving 15,000+ packages a day.
Infrastructure written down, reviewed and repeatable. Pipelines that ship on green and roll back without a meeting.
Cut system maintenance cost by 80% a year.
/work
Six cells. Each one is something running in production, or the open source that came out of building it.
Open a cell to read it
Mobile, web and infrastructure for a parcel locker network across the Balkans. 40 applications and services, 15,000+ packages moved a day.
Architecture and delivery for Austria's 5th-largest e-commerce app, adding $120,000 of client business value every six months.
Complex enterprise applications and POS integrations, plus the high-level refactors under them. 5+ new tools delivered at a 100% satisfaction rate.
A secure delivery pipeline built with a team of five: scanning, signed artifacts and reproducible environments. Maintenance cost down 80%.
A minimalist Node.js framework. TypeScript decorators for API definition, a custom dependency injection container and a route resolver.
An open-source module that injects Drizzle ORM into NestJS through modules and DI, with multi-connection management and a custom repository layer.
/stack
Chosen per problem, not per fashion. Everything here has been run in production by someone who had to fix it at 3am.
If your stack is not here, it is usually a week of ramp-up, not a reason to say no. Tell us what you run.
Ask us →/contact
Tell us what you are building and where it is stuck. You will get an engineer on the reply, not a sales sequence.