The Blueprint for Enterprise AI Transformation: Insights from Ben Kessler, CEO of 66degrees
The Blueprint for Enterprise AI Transformation: Insights from Ben Kessler, CEO of 66degrees
Successfully navigating the landscape of Artificial Intelligence is a top priority for enterprise organizations today. It’s not just about implementing a new technology; it’s a fundamental shift in how businesses operate, serve customers, and shape the future of work. Our CEO, Ben Kessler, recently shared his perspective on the blueprint for successful enterprise AI transformation, drawing on extensive experience helping companies modernize their data and infrastructure to prepare for AI transformation, build cutting-edge AI platforms and applications, and manage and scale their AI initiatives for sustained impact.
Ben’s core thesis is clear: successful transformation requires more than just technical prowess. It demands a strategic, long-term view of value, a deliberate approach to integrating AI across the organization, and a critical, unwavering drive from senior leadership.
Let’s explore the key elements of this blueprint.
Building a Robust Technical Foundation
AI solutions are only as strong as the foundation they’re built upon. Ben highlighted three critical areas for establishing this technical bedrock:
- Secure Cloud Infrastructure: A secure cloud environment is paramount. This isn’t just about protecting sensitive data; it’s also essential to prevent unwanted external influences on your AI models and the data they consume.
- Sufficient and Available Data: AI thrives on data. Ensuring an ample supply of data is crucial, but it must also be readily accessible. Latency in data availability—whether in data lakes, data warehouses, or through abstraction layers—can significantly hinder AI adoption. Having the right data in the right place, accessible when needed, is fundamental for enabling AI.
- Reimagined Business Processes: Simply plugging AI into existing workflows isn’t enough. Successful integration requires deeply considering and potentially rearchitecting current business processes. AI should integrate seamlessly, considering the overall organizational workflow and future direction, rather than merely replacing a single step. As Ben puts it, “To quickly recap, one the right infrastructure and securely set two the right amount of data and three the right business process and architecture for adopting AI.”
Rethinking ROI Beyond Initial Model Accuracy (“Yield”)
One of the most significant pitfalls in AI adoption is fixating solely on initial model accuracy, often referred to as “yield.” Companies sometimes abandon initiatives if a model doesn’t immediately hit a high accuracy target (e.g., expecting 95% but achieving only 65%).
Ben argues that this is shortsighted. He compares this to the early stages of transformative technologies like electric vehicles or achieving “5 9s” (99.999%) uptime for systems. Initial versions and attempts were rarely perfect but were necessary steps toward eventual success.
A holistic view of ROI is critical. Value and improvements emerge over time, even with lower initial accuracy. As we should never judge a book by its cover, we should not judge the success or failure of innovation by the first version. The key is to keep going. Continuing development, even with a 50%, 65%, or 75% yield today, will lead to experiencing ROI along the way and ultimately achieving the desired accuracy and full return on investment. As he states, “If your yield is 50, 65, or 75% today. Keep going. You will experience ROI and you will ultimately get to the yield that you want and experience that ROI along the way.”
