Traditionally, engineering teams spend significant time translating business problems into technical understanding before builds begin.
But in this case, detailed discovery work and a working prototype happened when a sales rep, someone who had a problem and Claude Code came together.
The problem? Virtual office onboarding is operationally complex.
Large operators need to verify company information, review director details, manage AML (anti-money laundering) checks and create auditable onboarding workflows for multiple locations and systems.
For many teams, that often means manual reviews, a bunch of different tools and slow onboarding processes.
Everything started with a prospect request
It all started when someone spoke to Matthias, one of our sales reps in France. They were looking for stronger AML and KYB (know your business) workflows.
Matthias started researching the problem and shortlisted two providers out of dozens based on pricing, capabilities, documentation quality, and API design.
Clear prompts and instructions came from a deep understanding of the problem in the context of Nexudus, our industry, and the person with the problem. Before long, Matthias built an initial prototype that connected two background check providers using Claude Code.
“What Matthias did was not simply prompt his way into generating some code to connect providers. He did a lot more than that.
He translated the prospect’s request into a system that could solve a real operational problem, and assembled a working end-to-end solution that could be demonstrated to a business.”
–Adrian Palacios, Co-founder and CTO at Nexudus
From prototype to production-ready
Security, scalability, multi-tenant architecture, maintainability and governance are critical; never more so than in workflows involving compliance checks.
“When Matthias presented the solution to the technical team for review, it honestly felt a bit magical.
The business understanding, research, experimentation, and workflow design had already happened before engineering formally entered the process.
Matthias had narrowed down providers, explored integrations and validated the customer journey before engineering began rebuilding the workflow properly inside Nexudus architecture.”
Adrian Palacios, Co-founder and CTO at Nexudus
From that point on, Engineering could focus on what they are uniquely good at:
- Making it secure
- Making it scalable
- Making it work reliably for thousands of operators simultaneously
- Making it fit into Nexudus architecture, standards, and infrastructure
The system was rebuilt inside Nexudus, integrating the tools needed to complete robust checks.
- Identity verification via Stripe Identity
- AML and sanctions checks via OpenSanctions
- Business verification workflows via Pappers
And all of this now means we have a workflow that allows operators to:
- Run AML and KYB checks during onboarding
- Verify company directors
- Review sanctions results
- Manually approve or reject flagged applications
- Rerun checks from the admin dashboard
- Manage onboarding workflows from a single system
The goal of this workflow is not to replace compliance teams or become a standalone RegTech platform.
Instead, it helps operators reduce manual admin and manage onboarding checks directly inside the systems they already use to run Virtual Office operations.
Features from the frontline
The most interesting part of this project may actually be what it says about how software teams are evolving.
Traditionally, identifying a customer problem, validating an approach, and building a working prototype would involve multiple handoffs between commercial and technical teams.
In this case, much of that discovery and validation happened before engineering formally entered the process.
A sales representative with a deep understanding of the customer problem was able to research providers, test workflows, evaluate integrations and build a working prototype that demonstrated the value of the solution.
Engineering then focused on the areas where their expertise matters most: architecture, security, scalability, governance and production readiness.
The result is not just a new AML and KYB workflow for Virtual Offices. It is an example of how AI is narrowing the gap between customer conversations and coworking software, enabling organizations to learn faster and deliver solutions with greater confidence.
