As spectrum is a limited and valuable resource, CSPs need to employ it as efficiently as possible. Yet many connected devices are either underutilizing or overutilizing the spectrum bandwidth available. This occurs because CSPs don’t have the necessary key performance indicators to assess how best to share different frequency bands among different connected devices. To help optimize the allocation of network resources, this Catalyst is training a large language model in performance evaluation. With access to 5G network data across all spectrum bands, the model will be able to measure the performance of network slices against service level agreements across a diverse range of scenarios and service categories. This AI-based solution will, for example, be able to monitor the network latency experienced by individual devices. The Catalyst plans to draw on and enhance existing network slicing APIs for product/service orchestration, as well as employing cloud-native applications based on a micro service architecture. This architecture will enable AI algorithms to dynamically share spectrum to ensure each device has the right amount of bandwidth, thereby enhancing the end-user experience. The Catalyst team believes TM Forum should bring in new best practice and standards for spectrum APIs for all 5G frequency bands. The solution will help CSPs better monetize 5G by efficiently utilizing all the available spectrum bands to support network slicing. Moreover, researchers in 5G networks, AI and network slicing will be able to use the datasets generated by the solution for in-depth analysis, model training, and performance evaluation, especially regarding resource allocation algorithms. Similarly, network architects and engineers will be able to use these datasets to consider diverse scenarios, aiding in the design and optimization of slicing strategies for different applications.