Your Best Workflow Engineer Probably Isn’t in IT

AI is turning operators into builders faster than enterprises expected.

Last week, we wrote about the rise of the "citizen developer" and how AI coding tools are enabling non-engineers to build internal software through natural language rather than traditional programming.

This week, the implications are becoming harder to ignore.

A private equity analyst can automate diligence workflows. A finance operator can create reporting tools that are directly tied to company systems. A customer success manager can build an internal AI assistant for support documentation.

Not prototypes. Not mockups. Actual working applications, automations, dashboards, copilots, and workflow agents.

The people closest to operational problems can now build solutions themselves.

For decades, building software within an enterprise required submitting requests to engineering teams, waiting through backlogs, and hoping your project would survive prioritization meetings. That assumption is starting to break.

AI is collapsing the distance between understanding a workflow and building software for it. And in many cases, the person who understands the workflow best is not an engineer.

This is the real shift happening underneath the "vibe coding" trend. The technology is impressive, but the organizational change is even bigger. Domain expertise is becoming as important, and sometimes more important, than technical expertise for first-generation internal applications.

The companies moving fastest are not treating AI-built applications as isolated experiments. They are recognizing that software development itself is increasingly distributed across the organization.

Engineering teams are still critical, but their roles are beginning to evolve. Instead of building every internal tool from scratch, engineering increasingly becomes the layer responsible for infrastructure, governance, security, integrations, and scale. Operators build closer to the business problem. Engineering ensures those systems can run safely inside the enterprise.

Finance teams did not wait for engineering to build Excel models. Marketing did not wait for IT to build reporting dashboards. Once software became accessible enough, the people closest to the work adopted it directly. AI is doing the same thing for internal applications on a much larger scale.

This creates real governance questions. As more employees build AI-powered tools and workflows, organizations inherit new challenges around security, permissions, compliance, data access, and maintenance. Many are already discovering shadow AI inside their organizations without formal oversight.

But the answer is not to stop it. It is to build the infrastructure that lets it run safely.

The organizations that figure out how to enable this new class of builder may end up moving dramatically faster than those still relying entirely on centralized development models.

The next generation of enterprise software will be built by the operators who understand the business best.

~Ryan Erickson, Founding Executive

About Shift
Shift helps organizations apply AI to real workflows through custom AI systems, workflow automation, and products like Basis and Greenlight, built for regulated environments and real-world adoption.

Greenlight
Helping IT teams give citizen developers a safe way to deploy and manage the apps and AI agents they build.

Basis
An AI intelligence platform for investment firms.