We've used Agent Builder, IXP, and Maestro for orchestration. We have created context-grounded searches or chatbots using context grounding, and we also use context grounding within an Agentic flow where a bot or agent reads ServiceNow tickets to determine whether it has been assigned to the right work group. If it has not been assigned correctly due to human error, it automatically reassigns it to the right group. I would assess Agentic AI's ability to extract data points as effective. We created an agent for the purpose of extracting data, though not in a document. We created what we call an email conversation agent that engages with a customer, vendor, or someone in an email conversation with a set goal of extracting specific data points for an action to be performed subsequently. The one automation we have deployed in production involves the agent interacting with team members to collect data points and create a ServiceNow ticket. We appreciate almost all features of UiPath Platform, with one of the most successful ones being the use of Coms Mining. Most of the new features related to AI have all been good, and we use Agentic. We have already created a few agents and have started to use Maestro for a few flows, though we are not yet there to orchestrate an end-to-end workflow. We have good ROI from the automations we have developed with UiPath Platform, which far exceeds the investment in the platform. It has saved us tens of thousands of hours in manual effort, making the team members happy to pursue other things of value.
