Choose the site problem
Select the risk, control behavior, test, integration, milestone, decision, or operating gap that matters.
Define the problem, the information and interfaces it depends on, the operating boundary, and how success will be evaluated. Then build a working system without forcing a full platform replacement.

Select the risk, control behavior, test, integration, milestone, decision, or operating gap that matters.
Identify the schedule, field, issue, cost, imagery, equipment, interface, test, or operating information required.
Document users, network, data, hosting, support, security, and optional AI requirements.
Deliver a usable software, controls, testing, or integration workflow that can be evaluated in the real environment.
Keep sources, assumptions, findings, and decisions accountable.
Expand integration, automation, or deployment only where value is demonstrated.

The technical pattern should reflect who operates the system, where compute and data live, how service access works, and what happens when connectivity is limited or unavailable.
Deployment can be scoped for connected, private, on-premises, edge, isolated-with-service, or fully disconnected environments. The proposed architecture must document connectivity, data movement, support access, and the controls required by the organization and project.

That creates a faster path to a working system and reduces the risk of procurement or security surprises later.