Turn Artificial Intelligence into Proprietary Intelligence

How to Win with AI

Decision 5: Operating Model

Decision 5: Operating Model

Design for speed, then build the culture to sustain it.

By Sarah Elk, Chuck Whitten, Hernan Saenz, Gene Rapoport, Nicolas Bloch,
Pascal Gautheron, and Anne Hoecker

Strategy and architecture mean nothing without an operating model built to deliver at pace.
  • Workflow redesign and workforce transformation must happen simultaneously; sequencing them kills momentum and allows the organization to revert to old patterns.
  • The operating model must be explicitly designed for speed: short delivery cycles, embedded feedback loops, structured release cadence, and decision rights pushed down to the squad level.
  • AI-native software development is the first function to fully transform and the most powerful proof point of what a five-times-speed model actually looks like in practice.
  • Broad access to AI tools matters as much as deep capability in a few teams. Identifying tinkerers, equipping them, and creating visible pathways from their discoveries to domain squads builds organizational pull that top-down mandates cannot.
  • Executives need their own individual contributor time. The tinkerer mindset extends to the top of the house, or it doesn’t extend at all. CEOs and their direct reports who personally build, experiment, and operate AI systems develop the intuition required to lead the transformation. Leaders who delegate the experience never close the gap between what they say and what they understand.
  • A desired outcome should be an organization where every nontechnical person has sufficient digital acumen to build in low- and no-code environments, demo possibilities to peers, and connect their own activity set to AI-driven value. It is a cultural transformation, not a training program.

Take a human-centric approach

The most common sequencing mistake in AI transformation is treating workflow redesign and workforce transformation as consecutive steps. Organizations redesign the process first, then figure out how to upskill or redeploy the people affected. By the time the workforce piece catches up, old habits have already taken hold in the redesigned workflow, and the momentum from launch has dissipated. The people who were supposed to work in the new way are still working in the old way because no one changed the systems, incentives, and role definitions that govern their day-to-day behavior.

The only way to avoid this result is to run tracks in parallel by redesigning the workflow and modernizing the workforce simultaneously, with the same leadership attention and the same sense of urgency. Take a human-centric approach to build a bright spot in the future way of working with a set of teams. This approach informs which solution should be scaled in real life and what upskilling is required for those roles and that specific change.

Speed is a strategic asset in AI transformation, not just a nice-to-have. The organizations that are pulling ahead are doing better work faster, which means they are learning faster and therefore compounding their advantage faster.

Speed is a strategic asset in AI transformation, not just a nice-to-have.

Building an operating model for speed requires four things:

  1. Short delivery cycles that force teams to ship something to the business every few weeks rather than building for months before revealing results
  2. Embedded feedback loops that capture signal from users continuously rather than through periodic surveys
  3. A structured release cadence that gives the business predictability about when changes that are ready to scale are coming
  4. Decision rights pushed far enough down the organization that squads can resolve most issues without escalating to a committee.

None of these things happens naturally in a large organization. They have to be designed in deliberately.

AI-native software development deserves special attention as a CEO priority because it is the single clearest proof point of what a transformed operating model actually looks like. When your software development teams operate with AI agents handling code generation, testing, and documentation, and they handle architecture, direction-setting, and the complex problems that agents cannot solve, the productivity difference is not incremental. Organizations that have made this shift report five to ten times the output from the same number of engineers. More importantly, the speed at which they can iterate on their AI applications across agents, workflows, and data pipelines is dramatically higher. Software development is the function most ripe for this transformation, and the organizations that do it first gain a compounding speed advantage in every subsequent domain they tackle.

How tinkerers become your superusers

Democratizing access to AI tools is equally important, and it operates through a different mechanism than top-down deployment. In every large organization, there are people outside of the obvious technology roles who are naturally curious about what AI can do, willing to experiment on their own time, and capable of discovering use cases and approaches that no central team would have thought to design. These are your tinkerers. The organizations that identify them, give them access to tools and sandboxes, and create clear pathways for their discoveries to reach the domain squads who can scale them are building a sensing network that multiplies the intelligence of their transformation program. These tinkerers become superusers who upskill everyone around them and show them the way to the future. This network is not a replacement for concentrated investment in priority domains, but rather a complement that keeps the program connected to the reality of how work gets done.

The tinkerer mindset must extend to the top of the house, or it will not extend at all. Don’t think it is about executives writing code. It’s about having spent enough hours using and building with AI to understand viscerally what it can and cannot do, where friction resides, what governance feels like in practice rather than in theory, and what your teams are actually being asked to do when you tell them to use AI. Without that experience, leaders make decisions about a technology they have only encountered through other people’s summaries.

The deeper challenge is time allocation. Most senior teams spend nearly all their hours running the business, and the change agenda gets whatever residue is left. AI transformation requires the opposite balance: meaningful, protected time blocked off to change the business rather than just run it, and a portion of that time spent inside the tools rather than supervising them from a distance. CEOs who delegate the experience never close the gap between what they say about AI and what they understand. The organization senses that gap, and the transformation slows.

The ultimate multiyear ambition is an organization where digital acumen is as common as financial literacy. Where every businessperson can build a working prototype in a low-code or no-code environment, demo it to their colleagues, and make a credible case for how it creates value. Where the gap between having an idea and being able to show it is measured in hours rather than months. Getting there requires a cultural transformation that goes well beyond training programs; it requires changing who gets hired, how performance is measured and celebrated, and what the organization treats as a core professional skill.

← Decision 4: Technology Architecture  Decision 6: The Learning System →

  • AI TRANSFORMATION: ENTERPRISE GLOSSARY

    The Foundation

    Agentic AI—An AI system that can plan and execute multistep tasks autonomously toward a defined goal, taking action on systems and data with minimal human intervention. The defining shift from generative AI (chatbots that answer questions) to agentic AI is the move from "tell me something" to "do something."

    AI Agent—A specific software entity built around a language model that perceives its environment, reasons about goals, and acts using tools (APIs, databases, other agents). Think of an agent as a digital worker with a defined scope, set of tools, and authority level.

    Frontier Models—The most capable AI models at the cutting edge of development (Claude, GPT, Gemini, and similar). They are powerful but expensive and change rapidly enough that any specific model choice is obsolete within months. Most enterprises consume these via API rather than building them.

    Probabilistic (vs. Deterministic)—A defining characteristic of AI systems: the same input can produce different outputs depending on context, memory, and model state. Unlike traditional software, which gives the same answer every time, AI is variable. This variability has major implications for testing, governance, and risk management.

    Stateful (vs. Stateless)—Agents maintain memory, context, and reasoning state across interactions rather than treating each interaction as independent. State persistence enables long-running workflows but also creates the risk that agent behavior drifts over time as the underlying data and models change.

    Compounding—The defining property of AI advantage that did not exist in prior technology waves: Data improves agents, agents improve people, people redesign work, and the redesigned work generates better data. The flywheel turns on its own. The gap between AI leaders and followers gets structurally harder to close every quarter, which is why early movers pull away faster than in any previous technology wave.

     

    The Infrastructure

    Enterprise Orchestration—The layer where a company manages its agents, its encoded know-how (skills, tools, and data ontology), and their access to data and systems as a single, governed enterprise asset rather than a scattered collection of vendor-supplied tools. It spans the registries that record what exists, the boundary at which agents access data and systems, and the controls that keep the estate inspectable and changeable. Distinct from agent orchestration, it is the narrower task of coordinating which agent does what within a single workflow.

    Harness—The execution infrastructure wrapped around an AI model that turns a probabilistic language model into a reliable working agent. Includes tool integration, memory management, safety controls, and observability. In 2026, the harness became the competitive frontier: The model is the engine, but the harness is the vehicle. Agent = Model + Harness.

    Multi-Agent System—A configuration where multiple AI agents work together on a workflow, each handling a specific subtask and passing results downstream. It is common in complex agentic work (e.g., a research agent feeding a writing agent, which then feeds a fact-checking agent). Multi-agent systems are where cascade risk lives.

    Model Context Protocol (MCP)—An open standard that defines how AI agents connect to external tools, data sources, and other systems in a consistent, governable way. Now stewarded by the Linux Foundation under the Agentic AI Foundation umbrella. Think of MCP as the USB-C of agentic AI: a common interface that lets any agent plug into any tool without bespoke integration work for each combination.

    Semantic Layer—A shared business vocabulary that defines, centrally and consistently, what your key business terms mean—revenue, customer, churn, margin, product—along with the rules and relationships between them. It sits between raw data and AI agents, so every agent reasons from the same definitions. Without it, every agent invents its own dialect of your business.

    AI Estate—The full inventory of AI agents, models, tools, integrations, and data pipelines operating across the enterprise. Managing the AI estate is the new infrastructure discipline, analogous to managing the IT estate or real estate portfolio. If you cannot inventory it, you cannot govern it.

    Shared Memory Layer—A persistent knowledge store that agents can read from and write to, allowing learning to compound across agents and over time. Without it, every new agent starts from scratch. With it, the tenth agent is faster and smarter on day one than the third agent was after six months.

     

    The Controls

    Registries (Agent/Tool/Skill)—Central records of every agent, tool, and skill operating in the enterprise, including purpose, owner, capabilities, and access permissions. If something is not registered, it does not run. The registry is the control surface that enables governance.

    Governed Gateway—The infrastructure layer through which every agent tool call routes, enabling policy enforcement, telemetry, and audit trails. Governance happens at this boundary, not inside the agent’s reasoning, which is technically inaccessible to inspection.

    Promotion Gates—Automated quality checks that determine whether an agent can move from development to production. Functions like quality control on a manufacturing line: the gate passes or blocks automatically, based on test results, threat assessments, and policy compliance.

    Evals (Evaluations)—The structured, continuous testing of AI agents and models to verify they produce accurate, safe, and useful outputs. Unlike traditional software testing, evals run continuously because AI behavior can drift as data shifts and underlying models update. Industry rule of thumb: Building the agent is roughly 20% of the work; evals are 60%; ongoing monitoring is the remaining 20% and never stops.

    Golden Test Suites—A defined set of test cases representing known scenarios that agents must pass before reaching production. Like a final exam, an agent has to retake it every time it or its underlying model changes.

    Shadow Mode—Running an agent in parallel with existing systems or human decision-makers without acting on its outputs. It is used to validate behavior before giving it real authority. Helps surface edge cases without putting business operations at risk.

     

    The Costs

    Tokens/Token Costs—The unit of compute that AI models consume to process and generate text. Every agent action consumes tokens, and at enterprise scale, aggregate token cost becomes a material line item that is invisible to traditional IT cost reporting. Token cost reporting is to AI operations what unit economics are to a manufacturing line.

    AI-Native Software Development—A development model where AI agents handle code generation, testing, and documentation while human engineers handle architecture and direction, achieving five to ten times the productivity of traditional development. The function most likely to transform first into an AI-led organization, and the clearest proof point of what an AI-transformed operating model looks like.

     

    The Risks

    Prompt Injection—An attack where malicious instructions are inserted into the input an AI agent processes, attempting to override the agent’s intended behavior. Analogous to social engineering against humans, but at machine scale and machine speed.

    Cascade Risk—The risk that an error or compromise in one agent propagates through downstream agents in a multi-agent chain, triggering a series of autonomous actions before anyone realizes something is wrong. One of the most underappreciated operational risks in enterprise agentic AI.

     

    The People and Governance

    Tinkerers—Employees outside traditional technology roles who naturally experiment with AI tools and discover use cases that no central team would have designed. A leader’s tinkerers are the organization’s sensing network for what’s actually working. Identifying and equipping them is a strategic move, not a culture program.

    Two Governance Motions (Run the Business/Change the Business)—The recognition that AI transformation requires a change-the-business governance motion (focused on learning, pivots, multiyear outcomes) running in parallel with existing run-the-business governance (focused on reliability, efficiency, milestone delivery). Running AI through existing governance alone is the single most common reason that transformations stall.

Quer saber mais?

Ajudamos líderes do mundo todo a lidar com desafios e oportunidades cruciais para suas organizações. Juntos, criamos mudanças e resultados duradouros.