Definitely, AI in a regulated industry is something that is a concern of the clients. However, Tulip can stick to 21 CFR Part 11 strictly when it comes to Tulip's AI capabilities. Tulip has already started to look forward to compliance-related things. For example, Tulip's AI widget offers data reading throughout the site. It does not allow any AI to interact with external systems when we apply firewalls on it. In this regard, if Tulip continues in that way, they can definitely stick to the governance and security. As of now, I didn't face any issue related to their AI capabilities. It is working fine but not accurate, which is a red flag in a GXP regulated system. This is my feedback on their AI capabilities. This has been introduced into the latest LTS version and needs more hands-on experience. Before that, I cannot give a direct justification or feedback on their AI capabilities being reliable and accurate. However, at the extent that I have used them, they are accurate and reliable as of now. Tulip is an emerging platform and being a cloud-native MES, it has a lot to improve. According to me, traditional MES like FactoryTalk PharmaSuite has more complex features and user-oriented use cases and complexities that the application itself handles. However, when it comes to Tulip, all the things that matter are in the hands of a developer. That is something that Tulip can improve on. Tulip can switch to a traditional way of delivering in the pharmaceutical industry while keeping the cloud-native impact intact. One improvement that I can give an example of is that Tulip has application templates. Instead of templates, they can offer concrete applications to their pharmaceutical or any specific industrial client for their specific needs, such as a recipe designer and production management for a pharmaceutical client. Then production or material tracking or BOM tracking for a discrete client. In that way, I can see an improvement where Tulip can perform well.
