Video Content
Advanced Steps in Tess AI's Agent Studio
Advanced steps are optional but powerful features inside Tess AI's AI Studio, that let you add more complex functionalities to your AI agents. They're super helpful whenever you need the AI to process info from different sources before generating an output.
Imagine, for example, you want to create an Agent that generates a product description from a PDF that has info about it. Just dropping the PDF as input isn't enough; the AI first needs to read and interpret the content from the PDF. That's when advanced steps come in.
Examples of advanced steps include:
PDF text extraction: Pulls raw text from a PDF;
Reading images with OCR (Optical Character Recognition): Extracts text from images in a PDF;
Reading selected PDF pages: Lets you choose which PDF pages the AI should process;
Web scraping: Pulls info from web pages. This can go from reading a specific article to collecting product data from an e-commerce site;
Google search: Does Google searches and uses the results as input for the AI.
You set up a sequence of actions: the user gives some input (like a PDF or a URL), the AI runs the corresponding step (for example, text extraction or web scraping), and then uses the info it got to generate the result, based on the prompt you set.
It's important to remember that every advanced step adds a step to the process, and that can make it take longer. So, it's key to use advanced steps only when you really need to, and pick the kind of step that fits your needs best.
The video also highlights how important it is to bring those advanced steps into the AI's prompt, so that it uses the processed info in a good way. In other words, it's not enough to just add an advanced step; you have to tell the AI how to use the data it gives you.