Governing and scaling
TL;DR: Running one good workflow across many teams depends on foundations that keep quality and risk in check as you scale.
- Put controls on the path the work travels, not at a gate after the fact, so they hold up under load.
- Make context retrievable so agents don't stall or guess.
- Verify output so a bad result gets caught before it counts as done.
- Let controls emit their own evidence as they run, so proof becomes a byproduct of the work.
A workflow that works once is not yet a practice you can run across many teams. Scaling multiplies not just useful work but also mistakes, so going from one workflow to many raises questions a single success didn’t. Early on, the question is what agents can do. As you scale, the harder and more useful question is how people and agents work together across many workflows at once. Staying in control of quality and risk at that scale depends on a set of foundations being in place before you expand.
Why foundations matter more as you scale
These foundations share a logic: put the control on the same path the work travels, rather than at a gate after the fact.12 Context that agents can retrieve keeps them from stalling or guessing. Verification catches a bad output before it counts as done. Controls that emit their own evidence as they run make proof a byproduct of the work rather than an audit-season scramble. Build these into the system rather than relying on instructions in a prompt. A limit that the workflow enforces automatically holds up under load, while a limit that depends on an agent following directions does not always survive contact with a new situation.
This is also how governance scales across teams that don't work the same way. When an organization spans many markets with different clients, compliance requirements, and developer constraints, requiring every team to use the same coding tools tends to stall, because the front end of the work genuinely differs from team to team. Consistent governance, context, and compliance in the delivery layer underneath, with each team choosing its own coding tools on top, puts the control on the path the work travels while letting the front end vary. That layer has to be shared for this to hold. When each tool brings its own policies and its own record of agent activity, governance splinters along the same lines as the tools. Offered as a reusable foundation, so teams don't each re-solve the same integration and governance problems, that arrangement is what turns one team's setup into something many can adopt.
As the pace increases, governance can’t stay a periodic checkpoint. It moves upstream into how systems are designed, shifting from gates that stop work to guardrails that shape it. Built this way, the same controls that hold quality and risk in check also let work move faster, because the checks travel with the work instead of waiting at the end of it.3
What governing and scaling involves
This topic covers three foundations:
- Make the codebase and its context agent-ready: Put the knowledge, decisions, and constraints agents need where they can retrieve them.
- Verify what agents produce: Use a layered check of deterministic tests, agent review, and selective human judgment so bad output gets caught.
- Own your controls and evidence: Design the controls your team is accountable for deliberately, and let systems produce evidence as they run.
Governing at scale also includes setting guardrails on what agents can reach and do, and measuring whether agents are actually helping rather than just staying busy. Both are dimensions this playbook will continue to develop.
How this connects to the rest of the playbook
Governing and scaling comes after Redesigning the work and Working with agents. The reusable pattern from reshaping a workflow and the checkpoints from working with agents are what the foundations here operate on. This is where a single success becomes something you can run at scale.
// References
- Hawthorne, Clare, and Ben Grinnell, James Nugent, and Dustin Warner. "Deliberate Compliance: A Field Guide to Owning Your IT Controls." Enterprise Technology Leadership Journal 3, no. 2 (Fall 2026). https://itrevolution.com/product/deliberate-compliance/.
- Cochran, Timothy, and Bryan Finster, Tim Ottersburg, Scott Prugh, and Amy Willard. "When Coding Is Free, Where Is the Bottleneck? A Guide for Leaders Building Agent-Ready Delivery Systems." Enterprise Technology Leadership Journal 3, no. 2 (Fall 2026). https://itrevolution.com/product/when-coding-is-free/.
- Eslick, Ian, and Collin Fox, Amanda Graham, Suzette Johnson, Chuck Lafferty, Tom Kilcommons, Rosalind Radcliffe, Cat Swetel, Annie Vella, Sam Yeats, and Robin Yeman. "The Sociotechnical Architecture of AI-Enabled Enterprises: Principles, Patterns, and Practices for Human-Governed AI." Enterprise Technology Leadership Journal 3, no. 2 (Fall 2026). https://itrevolution.com/product/sociotechnical-architecture-of-ai-enabled-enterprises/.
/ Go deeper on this topic
Governing and scaling
