Headwaters initiative to use Microsoft-Decision-1 to apply AI to corporate business decision-making

Headwaters and Headwaters Consulting announced that they will launch an initiative to apply AI to corporate business decision-making processes using Microsoft's decision-specific AI model, Microsoft-Decision-1. The initiative will target recurring decision-making tasks such as classification, routing, prioritisation and acceptance or rejection, with the aim of improving efficiency across entire business processes. Operation on Microsoft Foundry is envisaged as part of the implementation, and the policy is to accumulate the resulting knowledge as decision assets.

The initiative aims to optimise decision-making overall, including decision-making effort, processing time and AI usage costs, by structuring company-specific decision criteria and connecting decisions made by Decision-1 with AI agents and business systems. The accumulated knowledge is also expected to be deployed to other customers and business areas, enabling reuse through decision assets.

Against a market backdrop in which the scope of generative AI is expanding, verification and approval work can become a bottleneck in business decision-making. It has been noted that assigning all decisions to high-performance LLMs could increase costs and processing times. In this environment, the use of AI models specialised in selecting among defined options is attracting attention, and this initiative is one example of such an implementation.

Microsoft-Decision-1 evaluates predefined options and returns a decision together with probabilities. When a specified confidence level is exceeded, processing can continue, while a low confidence level can trigger a request for confirmation from another AI or a person in charge. From a security and governance perspective, an embedded design through Foundry is envisaged, with implementation support expected to be provided based on requirements such as authentication, authorisation and log monitoring.