This workshop aims to provide participants hands-on experience in creating smart workflows on a Model as a Service (MaaS) platform. Participants will learn how to assemble smart workflows that can react and respond according to the deployment of one or more generative AI models, customised datasets, and designated purposes. Participants will learn how to (1) enhance the workflow performance with their own datasets, (2) select appropriate generative AI models such as GPT-4 Turbo, Claude 3 Opus, and LLaMA 3, and (3) adopt other design options to optimise the created workflows. Additionally, the workshop will guide participants in publishing the workflows they designed as agents on web pages or application programming interfaces (APIs).