Custom Generative AI: The Next Productivity Frontier

In the 15th edition of this newsletter entitled “The 3D Decisions-Powered Enterprise Strategy for AI”, it was concluded that the “Enterprise AI Strategy” in one of its simplest images, can be seen as a “Series of Decision Points”. Each one of them can be powered by a “3-Dimension Decision” that should be taken or selected from the available “Eight Possible Options” for the “Value-added AI Models”. Furthermore, the 15th edition of this newsletter focused on the "Third Dimension" of the enterprise strategy for AI, which is the “Role” of the AI models in revolutionizing the “Enterprise Operational Model”. As mentioned, this AI model is either going to assist or augment the performance of certain tasks. Simply, it is either an “Assisting Intelligence Model” or an “Augmented Intelligence Model”.

However, this edition of this newsletter will extend the role played by the AI models to revolutionize the “Enterprise Marketing Model” too. Although each model contains three elements according to the well-known “Business Model Canvas”, the fundamental component of the enterprise operational model that the AI models are going to revolutionize is the “Enterprise Processes”. On the other hand, the fundamental component of the enterprise marketing model that the AI models are going to revolutionize is the “Customer Relationships” and how the enterprise is managing them.

It is worth mentioning that the overall “Enterprise Business Model” consists mainly of the operational and marketing models related to the initially intended value proposition plus the "Financial Model". Furthermore, it is worth highlighting that this financial model, as the third and final component of the business model, is like the body temperature. I can't be directly improved and to keep it in good shape the body should be kept overall healthy in the first place. Improving the financial model can be only achieved as a result of the improved operational and marketing models. 

Here comes the value of the "Customized Generative AI" value-added models that can act as assisted or augmented intelligence to improve enterprise processes and customer relationships. Two of the most evident examples are using Generative AI to internally support the call agents instead of second-line support as the first resort and/or the customer-facing chatbots.

The basic and most fundamental building block of these customized Generative AI models is the “Foundation Models”. Simply, the foundation models are "Pre-trained Open-Source Large Language Models” that were trained on a massive amount of linguistic data sets and already consumed an enormous amount of resources to be trained in terms of infrastructure, time, and cost. These models can be fine-tuned using the “Enterprise Data Assets” to capture “Enterprise Knowledge” and be able to internally support the enterprise.

However, here comes the challenges. The “Critical Success Factors” in this situation will be selecting the best-fit model in terms of required training resources, the total cost of ownership, and the minimum size of the training datasets required to take off the fine-tuning training task. Not easy tasks at all, and they are going to differentiate creative enterprises from the crowd.

Hence, to conclude, "Customized Generative AI" models can revolutionize both "Enterprise Operational and Marketing Models", and push the current productive levels to the next frontier by carefully selecting the "Best-Fit Foundation Models" that can turn the "Enterprise Data Assets" into "Actionable Knowledge" while carefully taking into consideration the previously mentioned "Critical Success Factors".