Due to Automic Workload Automation's proprietary scripting language, upgrading it can be extremely challenging, unlike other workflow automation products that offer seamless migration. This inflexibility makes Automic Workload Automation the most complex and restrictive solution in the market. Choosing Automic Workload Automation essentially locks us into their ecosystem, making it nearly impossible to switch to a different product. Therefore, I strongly advise against using Automic Workload Automation. Automic Workload Automation's AI capabilities are limited. Most traditional workflow products lack robust support for AI workflows. Airflow might be a suitable option for AI workflows. However, if real-time AI processing is required, a different product altogether is necessary. For example, in the field of genetics, if a workflow involves thousands of jobs, traditional workflow products such as Automic Workload Automation may struggle to handle such a large workload. The maximum capacity of these products might be around 1,000 or 2,000 jobs. In contrast, a genetic workflow could involve up to 100,000 jobs, requiring a completely different workflow product specifically designed for such large-scale processing. Mainstream workflow products like Automic Workload Automation offer similar functionalities and are widely used around the globe. These products typically check for process completion every second. However, in high-performance computing and emerging fields like medicine or ophthalmology, we need to control thousands of jobs simultaneously, requiring millisecond-level process completion checks. To achieve this, we can store event data in databases or perform on-the-fly checks. Additionally, we need to integrate workflow control with workload management to prevent machine overload. These requirements make it unsuitable for tasks like controlling genomic workflows.
