Klaria: Sales & Operations Platform
An operations platform for a Norwegian cleaning company: sales pipeline, partner network, job dispatch and invoicing, with the infrastructure and observability underneath it.
Klaria is a Norwegian cleaning company that sells jobs and subcontracts the work to partner firms. The platform runs that cycle end to end: lead intake, a sales pipeline with a calling queue, quotes and pricing, dispatch to a partner, completion, and invoicing that lines up with the accounts.
Underneath it sits a second-generation build on ASP.NET Core and Postgres, deployed as rootless Podman containers on Hetzner machines behind Traefik, with OpenTelemetry into a self-hosted SigNoz and nightly backups to a separate provider. Reporting comes from database views, so every screen quotes the same number. The engagement began as a freelance project and became an ongoing one.
Case Study
The Problem
Klaria sells cleaning jobs and subcontracts the work to a network of partner firms. The business was growing fast, and it had reached the point where a growing team needed one system covering the whole cycle rather than several that each held part of it. That meant the full path from first call to paid invoice in one place, with reporting the owners could plan against and an operations view that keeps a busy day legible.
The Approach
I put the reporting on database views so every screen reads the same definition, and tied it to the accounting system so the monthly figures line up with the books. Then the pieces the daily work needed: a calling queue that surfaces only what can be acted on now, an operations view for the day's exceptions, partner onboarding with a signed agreement step, reporting per broker chain and office, and a message catalogue where customer-facing text is data the owners edit themselves rather than markup buried in code. The second-generation platform is ASP.NET Core with its own Postgres and a React front end, running as rootless Podman containers under systemd on Hetzner machines behind Traefik. Everything emits OpenTelemetry into a self-hosted SigNoz, and the dashboards are generated from code rather than clicked together in a UI.
The Outcome
The platform runs the business day to day, and the figures reconcile against the accounts. Backups go nightly to object storage at a separate provider under a retention lock, and staging rebuilds itself from the production dump every night, so the restore path gets exercised daily instead of for the first time in an emergency. What started as a freelance build turned into something longer: Klaria took me on, and I have kept building it around a full-time job.
“Helgi is an exceptional developer, highly skilled, incredibly responsive, and someone I fully trust. He doesn't just build what you ask for; he challenges ideas, thinks ahead, suggests better solutions, and adapts to what you actually need. What really stands out is that he genuinely cares about helping your business succeed, not just getting paid.”
Klaria AS