Build
Enabling technical and non-technical people to make things with AI through supported tools, access, model keys, and data.
Build covers how people make AI applications and automations at work. It includes their tools, identity, model credentials, data access, and route from an idea to a running system.
Build starts with the sanctioned path people use, whether that is a ready-made assistant, a low-code workflow, a supplier application, an API, or custom development. The starting enterprise AI guide applies all five BROCS scopes to each entry route.
A coding-assistant purchase covers one tool. The applications people create still need an organizational path into production.
Coverage
Without Build, AI work remains ad hoc. Covering Build alone leaves applications without a managed runtime and provider keys without attribution. Teams also accumulate useful tools that have no owner or operating path.
Operating tests
- Measure the time from a new person's start date to their first shipped AI change, separately for engineers and non-engineers.
- List everything that would need rebuilding if the primary AI tool changed next quarter. Identity, keys, data access, and runtime should remain independent of that tool.
Tradeoffs
Build needs enough Secure and Control safeguards to protect the organization without making the sanctioned route impractical. When approval and access take longer than an unsanctioned option, people have a strong incentive to work around them. Platform teams can address that by making the supported route fast and predictable.
In this capability
- Tooling: The shared enablement layer beneath IDEs, agent CLIs, notebooks, and chat, designed for changing tools.
- Access: Identity for builders: who gets which AI tools, how they authenticate, and how a new hire is productive on day one.
- Keys and models: Distributing, scoping, rotating, and metering model credentials. BYOK, local models, models through your own cloud, and BAA-covered endpoints.
- Data access: Connecting builders to the data they need through masking, scoping, approvals, and permission-aware retrieval.
- Idea to production: Golden paths, templates, and scaffolds. DORA lead time applied to AI work, and what it takes for non-engineers to build like engineers.