The product was there and had been built pragmatically: a live platform, real users, AI features already in production. What did not hold was the environment underneath — designed to launch, not to absorb growing traffic and concurrent users.
The risk in these situations arrives later than today's outage. It shows up when every increase in usage forces a redesign, while nobody can say where the data goes or on what terms the service is provided.
Infrastructure
Migration onto something that scalesExporting the platform from its existing environment and configuring application, database, access, backups and deployment properly. The goal is absorbing growth without redesigning the system at every jump in usage.
Observability
Monitoring from day oneMonitoring set up before the first incident. Knowing how the system is doing is as much part of the infrastructure as the servers.
Process
Faster development, better controlledOrganising the environment and the development process — repository, review, coding assistants — to move faster along a documented, traceable path.
Assessment
Data, AI and external vendorsA technical and organisational review of the AI features and data flows: what leaves the perimeter, towards which providers, and what that implies for GDPR, the AI Act and security.
Terms
Conditions of use and licensingTerms of use and licensing follow from the assessment, matching how the platform actually works. Ordinary matters are handled in-house; legal review is kept for the questions that deserve it.
From outside nothing shows. No new features, no redesign. What changes is the trajectory — the platform can grow without being rebuilt, the team ships faster with more control, and the questions about data and terms have a written answer before anyone turns up to ask them.