Checking Status of Agentic Solution Evaluation

​@jhonnie rafael perdomo You did not pass. You’re using your AI Agents as AI Skills. AI Agents must have autonomy to choose its own workflow and when it uses its tools (or doesn’t use them!). 

I already fixed it a few days ago—I used skills to make the agents autonomous—but the link you sent me for re-uploading isn't working; it says I've already uploaded it, but I haven't. Here is the link:https://upskill.automationanywhere.com/ai-automation-engineer-assessment-resubmission

Please review it again, as my project is currently in the check-in stage.

​@jhonnie rafael perdomo I will put it back in the queue, but if the message you saw said, “you already submitted” instead of “you already resubmitted”, you were using the wrong link.

Exactly. If you could please help get it reviewed so I can upload it—perhaps with a new link?

​@jhonnie rafael perdomo I just reviewed your submission again by putting it back in the queue manually. You are really using your three AI Agents as if they were AI Skills. This means the workflow of the AI Agents isn’t really autonomous. It doesn’t help that the AI Agents only have a handful of tools each. 

Remember that not all use cases require an AI Agent. You could solve your use case using a Process and some AI Skills. Nothing really requires the use of an AI Agent. Even human-in-the-loop could be accomplished in a Process based on the logic you have in place.

So you're telling me to use just one agent with the abilities I have?

​@jhonnie rafael perdomo What I’m saying is that you need to take a close look at the business use case you chose. It may not need an AI Agent.

Think about how a Process is logic-based. It executes elements and can use an IF THEN to branch to other elements.

Now think about an AI Agent as receiving some data, using that data to determine its own process autonomously and its tools autonomously. It’s not logic-based, it is allowing the AI Agent decide what to do and when to do it.