Inside Aethon's AI Transformation Initiative: Early Results from the Portfolio

In October 2025, Aethon Capital announced the launch of a portfolio-wide AI Transformation Initiative supported by a dedicated AI Operations Center and $200 million in centralized investment. Six weeks in, the early results are exceeding our expectations—and they are also teaching us valuable lessons about what it really takes to capture AI productivity at scale.

In this note, I want to share what is working, what is harder than expected, and how we are evolving the program based on what we are learning.

**What is working: high-leverage, well-defined use cases**

The most successful early pilots have focused on use cases that share three characteristics: clearly defined inputs and outputs, high transaction volume, and measurable cost or quality outcomes. Three categories stand out.

**Customer service and support**: Across four portfolio companies in different sectors, AI-powered customer support agents are now handling 35-55% of initial customer inquiries without human intervention, with customer satisfaction scores meeting or exceeding human-handled equivalents. The cost per interaction has fallen by an average of 60%, and human agents are now able to focus on complex cases where their judgment and empathy add the most value.

**Software engineering productivity**: At our software portfolio companies, the rollout of AI coding assistants has improved engineering throughput by 25-35% for routine development tasks, with even larger gains in specific areas like test generation, documentation, and code review. Importantly, code quality has not deteriorated—and in some metrics has improved—reflecting the value of AI-assisted review and testing workflows.

**Sales operations and intelligence**: AI-powered prospecting, account research, and meeting preparation tools are significantly compressing the time sales representatives spend on administrative work. Across pilot deployments, sales reps are spending 30-40% more time in actual customer conversations, translating to measurable improvements in pipeline coverage and conversion rates.

**What is harder than expected: change management and data**

The technology side of AI deployment is increasingly straightforward. The harder challenges are organizational and informational.

On the organizational side, AI adoption requires meaningful change management—training, workflow redesign, role redefinition, and cultural reinforcement. The portfolio companies that are seeing the strongest results are those whose CEOs are personally engaged and visibly committed to the transformation. Where leadership is delegating AI to a peripheral function, results have been muted.

On the informational side, the quality, accessibility, and governance of internal data has emerged as the binding constraint at many portfolio companies. AI agents are only as good as the data they can access, and most companies have meaningful work to do to make their data discoverable, structured, and trustworthy. We are increasingly investing in data infrastructure as a precursor to AI deployment.

**What we are learning: the need for centralized capability**

Perhaps the most important lesson of the past six weeks is the value of centralized AI capability across the portfolio. Individual portfolio companies—even well-resourced ones—often lack the specialized expertise, vendor relationships, and learning networks needed to deploy AI effectively. By centralizing infrastructure, expertise, and playbooks at the Aethon AI Operations Center, we are accelerating adoption across the entire portfolio while avoiding redundant investment.

We have also learned the value of cross-portfolio knowledge sharing. The lessons one portfolio company learns from a customer service deployment are directly applicable to others. Our quarterly AI summits, which bring together CTOs and AI leaders from across the portfolio, have become surprisingly high-value events—not because of any presentations, but because of the peer learning that happens in the hallway and over dinner.

**Looking ahead**

The next phase of the AI Transformation Initiative will focus on three priorities. First, scaling the most successful early pilots from individual use cases to enterprise-wide deployment. Second, addressing the data infrastructure constraints that are limiting AI value capture at many portfolio companies. Third, deepening the integration of AI capability into Aethon's investment underwriting process, both as a value creation lever and as a risk consideration.

We are also evolving our hiring profile across the portfolio. The companies that succeed with AI will be those whose leadership combines deep functional expertise with genuine curiosity about how AI is reshaping their industry. We are increasingly looking for that combination in CEO, COO, and Chief Product Officer searches.

AI is the most consequential technology shift of our generation, and the value creation opportunity for companies that adopt it well is enormous. The early results from Aethon's portfolio reinforce our conviction that disciplined, well-resourced AI deployment can be a significant driver of EBITDA growth and competitive advantage. We are still in the early innings, but the trajectory is clear.