From Disconnected Operations to Complete Business Visibility

Customer logo
Case study thumbnail
Remote Talents is a Netherlands-based recruitment agency that places remote talent with international clients. Sales, recruitment, account management and finance each worked in their own tools, and the numbers rarely matched. We gave the agency one reporting platform that every department now relies on.

Founded in

Netherlands

Became Customer

Website

About Remote Talents

Remote Talents is a recruitment agency based in the Netherlands. It helps companies hire skilled remote professionals and supports them after the placement through account management. The business covers the full cycle: generating leads, selling the service, recruiting candidates, managing the client relationship and invoicing.

Each of those steps is handled by a different team, and each team had chosen the tools that suited its work best. That made sense for the teams, but it made the business very hard to read as a whole.

The challenge

Leads came in from ads and cold email. Setters and closers worked them in the CRM and the phone system. Recruiters filled vacancies in an applicant-tracking system. Account managers followed placements, and finance invoiced clients and reconciled bank statements in accounting software. In total, nine source systems held a piece of the story.

Every one of those systems produced its own reports, and none of them agreed. Management meetings often started with reconciling figures instead of making decisions. Closers typed their end-of-day results into a spreadsheet by hand. An early Power BI model existed, but its measures lived in one file, refresh limits capped how fresh the numbers could be, and nobody could tell whether a dip was a bad week or a broken data feed.

Where the numbers broke down
  • Each tool had its own idea of a month, an owner and a closed deal, so the same question gave different answers depending on where you looked.

  • Ad accounts were billed in AED, bank accounts held AED, USD and GBP, and invoices were issued in EUR, which made revenue and cost comparisons unreliable.

  • When a third-party integration created a new sales pipeline and reassigned owners on about 500 deals, months that had already been reported suddenly changed.

  • Some data was being lost without anyone noticing. A feed from the recruitment system looked healthy but had been missing 35 of 334 jobs for weeks.

Our approach: business decisions before dashboards

We did not start by building charts. We started by agreeing with each department lead what every number should mean. What counts as a close? Which date decides the month a deal belongs to? Who gets credit when a lead changes hands? Each open question went to the owner as a short decision with a recommendation, and was built as soon as it was answered. 31 KPI decisions were documented this way, and that decision log is now the reference for how the agency measures itself.

Before switching anything off, we translated every measure from the old Power BI model into the new platform and checked it against the previous values, so the team could see exactly where and why a number changed.

What we put in place
  • A 7-page official report, one page per department plus an overview and data quality, with around 216 tiles. Every KPI shows its definition one hover away.

  • A self-service editor where team leads duplicate any page and rearrange it for their own needs, with an AI assistant that drafts new charts from a sentence.

  • An AI chat that answers questions in plain language and saves useful answers as shareable insights.

  • Smart morning alerts when a number departs from what is usual, with thresholds tested against history and held back when a feed is merely late.

  • Google Sheets both ways: scheduled exports into the sheets people choose, and team sheets brought in as governed data after one approval.

  • An open report API with personal keys, so data can be pulled into Excel, Google Sheets or Python with the same access rules as the app.

  • Access by department, with finance visible to management only.

Under the hood: a modern data warehouse

Solverix OS for Remote Talents is built as a full data platform. Data from the nine sources is loaded into a cloud data warehouse on a schedule, then transformed in clear layers: raw data as delivered, cleaned and typed staging tables, and 45 business-ready tables and dimensions that every report reads from. This ELT approach means raw data is never lost, every number can be traced back to its source, and all currencies are converted to EUR at official monthly rates.

On top of the warehouse sits a semantic layer with 240 KPIs, each defined once with its meaning, calculation, date basis and caveats. The report, the AI assistant, the exports and the API all use that same definition, so two screens can never disagree. Data is typically about an hour old, and every page shows exactly when it was last refreshed.

Around 550 automated tests run every night. If a check fails, the platform keeps the last good numbers on screen instead of publishing a wrong one, and alerts the team with the likely cause. Guards also fail the moment a source introduces a pipeline or stage the reports do not know yet, so changes in other tools are caught the next morning, not at month end.

Why not Power BI or Microsoft Fabric alone?

Power BI is a good visualization tool, and Microsoft Fabric adds storage and pipelines around it. But neither one decides what a close is, keeps 240 definitions consistent across report, AI and API, or notices that a connector silently skipped jobs. With a packaged BI stack, that work still has to be designed, built and maintained by someone. Remote Talents had already seen the limits: logic locked in one file, refresh caps and no way to tell a data problem from a business problem.

The custom approach let us keep the agency's business rules as code, test them every night, run on one server plus a pay-per-query warehouse, and avoid per-user licensing and vendor lock-in. The team still gets familiar, interactive reports, with a data foundation it can trust underneath.

Built for the team to keep growing
  • Department leads build their own views and ask their own questions, without waiting for a developer.

  • Admins can refresh a source, rerun a load, rebuild the reports or snooze a known issue from inside the app, without server access.

  • New definitions and requests go through regular review sessions with the Solverix team and become tracked items with a release number. The platform shipped 19 releases in its first five weeks after go-live.

  • Online support and an operator guide, runbooks and a generated KPI glossary make the platform easy to hand over and maintain.

The result

The platform went live on 23 September 2026 and the owner retired the Power BI prototype the next day. Every department now works from the same figures, and meetings start with decisions rather than reconciliation. Manual end-of-day reporting is gone, because results are read directly from CRM stage changes. Data problems surface with their cause before anyone acts on a wrong number, and running costs stay low.

Closing perspective

For Remote Talents, the hardest part was never the technology. It was agreeing, as a business, on what the numbers mean. Once those decisions were written down, a proper data warehouse and semantic layer enforced them everywhere, and the agency gained one version of the truth that every team accepts.

Solverix understood our business faster than we expected. They connected the dots across our teams, simplified our reporting, and brought some seriously smart ideas to the table. Great technology, great people, and a real impact on how we operate.

Roy Lammers

CEO , Remote Talents

On this page

Keep exploring how modern teams ship

Keep exploring how modern teams ship