Ask most enterprises how their AI is going and you'll hear about projects. A chatbot here, a copilot there, a pilot in one department, a proof-of-concept in another. Each has a start date, a budget, and a definition of done. And that, more than any shortcoming of the models, is why so many of them quietly fail.
The numbers are stark. Independent research from the RAND Corporation puts the failure rate of enterprise AI projects above 80% — roughly twice that of conventional IT projects. MIT found that 95% of enterprise generative-AI pilots deliver no measurable impact on the bottom line. The common thread is not the technology. It is the mental model: organisations keep treating AI as something you install, when it is something you become.
The problem with "project" thinking
A project has edges. You scope it, ship it, and move on. That shape works for a payroll system or a CRM migration — the kind of change that swaps one tool for another without touching how people actually work. AI is not that kind of change. It alters what a role is for, how a process flows, and who — or what — does the work. Bolt it onto an unchanged operation and, as McKinsey puts it, you get a slightly faster version of a broken workflow.
Project thinking also has an expiry date built in. When the initiative "ends", so does the learning, the governance, and the momentum. The pilot that dazzled in the demo never adapts to the messy reality of production. This is why Gartner expects a third of generative-AI projects — and more than 40% of the newer agentic ones — to be abandoned or cancelled. They were built to finish, not to endure.
What an operating layer is
An operating layer is not a system you buy; it is a capability the organisation runs on. It has four properties a project never will:
- →Durable. It outlives any single use-case. New workflows plug into a foundation that already exists, rather than starting from zero each time.
- →Hybrid. Human employees and Digital Employees work side by side as the default way work gets done — not as a novelty bolted on the side.
- →Governed. Clear boundaries on what each agent may do and touch, auditability for every action, continuous evaluation in production, and a named human accountable for the outcome.
- →Improving. Every deployment makes the next one faster — capability, data and trust compound instead of resetting.
That last property is the quiet superpower. Projects depreciate; an operating layer appreciates. The gap between the 5% who win with AI and the 95% who don't is not a gap in technology — it is the difference between organisations that transformed and organisations that installed.
It's a journey, and you can locate yourself on it
No organisation jumps straight to an operating layer. There is a curve, and value compounds only toward the far end of it:
Experimenting — scattered, IT-led pilots and impressive demos, but no P&L impact. Adopting — workflows redesigned in a few functions, people reskilled, the first real outcomes. Scaling — governed rollout across functions, Digital Employees alongside people. Operating layer — AI is simply how the organisation runs. Most enterprises are stuck at the first stage. Knowing which stage you're in — and what the next one demands of your people and processes, not just your platform — is the most useful strategic question a leadership team can ask this year.
Why local depth decides it
An operating layer is, by definition, built on your operation — your processes, your data, your people, your regulatory context. That is why generic, global, install-and-hope approaches stall, and why the work has to happen with local depth: inside the business, with the teams who will run it. For organisations in Singapore and across ASEAN, that means a partner who works in your context and stays until the new way of working ships and your people own it — not one who hands over a licence and a slide deck.
This is the work PPIC Technologies does. We help organisations move up the maturity curve — reorganising the workforce through training, redefining the processes through implementation-led mentorship, and bringing governed Digital Employees into the operating model alongside their people. Not another AI project. The layer your organisation runs on.
Ready to build your operating layer?
We help organisations transform the workforce, redesign the processes, and bring governed Digital Employees into how they run — locally, and with you.