From AI Curiosity to Mission Capacity

How Flatter built an Internal AI Champions Model to Accelerate Innovation, Develop Its Workforce, and Prepare Leaders for an AI Enabled National Security Environment

The Challenge: The Technology Was Moving Faster Than the Workforce

In 2024, Flatter, Inc. began confronting a problem that would soon become familiar across the national security enterprise: artificial intelligence was advancing rapidly, but access to increasingly capable technology did not mean that a workforce was prepared to use it.

For Flatter, the stakes extended beyond productivity.

As a federal contractor supporting national security missions and senior leader development, the company viewed AI literacy as part of a larger question of readiness. CEO Brittany Chiang describes the original concern in competitive terms: the United States would need to capture the benefits of AI quickly enough to maintain advantages not only against strategic competitors, but within a rapidly changing defense technology environment.

At the same time, Flatter anticipated that adoption would face two very human barriers—resistance to change and a substantial learning curve.

Instead of “Which AI tool should we buy?” Flatter began asking: “How do we develop people, our most precious resource, to responsibly lead through this unprecedented period in history — to navigate change, and translate emerging technology into mission capacity?”

Starting with Experimentation, Not an Enterprise Rollout

The result became AÇAI: AI Champions for Accelerating Innovation, an internal workforce-development and adoption model built around AI literacy, experimentation, change leadership, coaching, and practical application.

Its underlying premise is straightforward: AI transformation is ultimately a workforce transformation.

AÇAI did not begin as a polished curriculum. It began as an experiment.

According to Flatter CEO, Brittany Chiang, Flatter initially placed AI licenses in the hands of a small group of employees and executives and encouraged them to explore. The company then established recurring workforce-development sessions and invited practitioners with dramatically different perspectives on AI—from Hollywood animation to defense applications involving target identification. Other sessions focused on immediately accessible work, such as using generative AI to prepare for meetings more efficiently.

This mattered because Flatter was deliberately developing familiarity before prescribing solutions. Employees could see AI operating in different environments, experiment with tools themselves, compare experiences with colleagues, and begin developing their own judgment about where the technology was useful.

Over time, an important insight emerged:

Good AI users were not necessarily the people who knew the most about AI. They were often the people who knew how to ask better questions.

That observation changed the direction of the program.

Flatter was already an International Coaching Federation coach-education provider. Coaching and AI appeared, at first glance, to be unrelated disciplines. In practice, Flatter saw considerable overlap. Both depend upon inquiry. Both require clarity about the desired outcome before jumping to a solution. Both benefit from surfacing assumptions and context. And both become dangerous when the person providing assistance takes ownership of a decision that properly belongs to the leader.

Flatter therefore began combining AI literacy with coaching capability.

The AI Champion was no longer envisioned simply as the employee who knew the newest tools. The Champion could become the person capable of helping another leader determine whether, where, why, and how AI should support the work.

A People-First Model for AI Adoption

The resulting AÇAI model deliberately rejects the idea of trying to “insert AI everywhere.” Its program materials instead define AÇAI as a workforce-development effort intended to move organizations from AI curiosity toward responsible, effective, and scalable adoption by preparing designated Champions to integrate AI into real workflows, support change management, and maintain ethical standards.

Champions are expected to become advocates and experimenters within their teams, identify real problems that might benefit from AI, pilot use cases, continue developing their AI fluency, and share learning across organizational boundaries. Flatter supports that work with interactive sessions, coaching, access to tools and prompts, guest experts, and a community of other change agents.

The distinction is significant. Traditional enterprise technology training often begins with a system: Here is the software. Here are its features. Here is how you use it. AÇAI begins with the work.

The current curriculum explicitly instructs participants to “begin with the leader’s work, not the tool.” Students establish what the leader wants to accomplish, understand context and constraints, ask questions that create awareness, and introduce AI only when it is authorized, appropriate, and useful to the leader’s objective. That principle is captured in an internal phrase that Chiang identifies with the program:

“Inquiry first, mission first, with a leader in the driver’s seat”

Flatter is an ICF coach-education provider. That experience led to an unconventional insight: coaching may be one of the missing competencies in enterprise AI adoption.

A good AI Champion shouldn’t simply arrive with an answer.

They should be able to help a leader understand the desired outcome, examine the current workflow, surface friction and constraints, determine what information can responsibly be shared, consider risk and controls, and identify the smallest useful experiment.

The current AÇAI curriculum has grown into 60 instructional hours—30 live and 30 asynchronous—combining ICF Level 1-aligned coaching education with applied AI practice.

Participants move from responsible AI use and coaching fundamentals into workflow analysis, automation opportunity spotting, human-AI engineering, knowledge readiness, change leadership, autonomy, governance, pilot scaling and stakeholder adoption.

Delayed Generation — Plan Before You Prompt

Generative AI makes producing an answer almost effortless. AÇAI teaches people not to start there.

Students learn to understand the work before generating the solution. The program’s nine-block Solution Coaching Canvas moves learners through this planning process. Built on the foundation of the Project Management Institute’s CPMAI certificate program, Business Planning and Data Planning are two of the most important phases to beginning an AI Project.

The competitive advantage of AI will not come simply from giving everyone access to the same frontier models.

It will come from developing a workforce capable of asking:

That is the workforce AÇAI is designed to develop.

AÇAI isn’t lecture-only training. Every instructional hour requires practice, application, reflection, or a tangible artifact. Participants coach, observe, experiment, analyze workflows and work through realistic cases.

PersonaOPPs provides AI-enabled role-play for selected scenarios so learners can practice consequential conversations in a controlled environment.

What’s Next

The current program combines ICF Level 1-aligned coaching education with applied AI practice and uses PersonaOPPs in selected exercises.

Flatter’s roadmap is to pursue the next evolution of the model: submission to the International Coaching Federation for recognition as a Level 1 AI specialty coaching program.

The ambition is larger than teaching people to use AI. It is about creating organizational capacity that will influence national security.