Enabling new use cases for decision models beyond high-volume cheap decisions (www.uprightproject.com)

Upright has launched Eevo, a consequential decision model that expands the use cases of decision models beyond just high-volume, cost-effective decisions. Unlike traditional generative large language models (LLMs), decision models like Eevo operate by selecting options from a predefined set and generating associated probabilities, making them quick and inexpensive. Initially positioned for high-volume applications, Eevo’s unique capability lies in enabling AI agents to make nuanced decisions by utilizing its efficient API for assessing consequences and expected values. This efficiency allows agents equipped with advanced models to leverage Eevo for critical decision-making without incurring prohibitive costs. The introduction of Eevo is significant for the AI/ML community, as it challenges the notion that decision models must compromise on quality for speed and cost. While Eevo may not match the output quality of leading generative models, its training incorporates various reasoning skills and comparative assessment techniques that enhance its versatility, enabling it to excel in complex decision-making tasks. Furthermore, the model's ability to operate on-premise offers organizations greater control over their AI infrastructure, paving the way for customized decision models tailored to specific needs. As a result, Eevo represents a promising advancement in the practical application of decision models in AI.