E2. Adding Content to AI Twins
An AI Twin's core grounding comes from the same three layers described in IV. Insights & Intelligence: the Behavior, Intent, and Mentions reports built into its persona. You can confirm exactly what that foundation contains by opening a Twin's profile card before a conversation. But research rarely stops at what a platform already knows. Sometimes the material that matters most is a brand guideline, a customer satisfaction file, or a document your team is holding that the Twin hasn't seen yet, and that's what the Assets tab is for.
The Assets Tab
Sitting next to your conversation, the Assets tab holds everything grounding the current Twin: the underlying intelligence reports powering the persona, alongside any brand guidelines, satisfaction metrics, or data files you've added yourself. It's the place to check, at a glance, everything a Twin's answers in this session are allowed to draw from.
Adding Your Own Material
Click the plus icon inside Assets to add to that foundation. You can attach image files, PDFs, or website URLs, or run a quick search across the contents of the current chat itself. Once added, that material becomes part of the context available to the Twin for the rest of the conversation, alongside its built-in report grounding.
This matters most when your research question depends on something specific to your organization rather than general audience behavior, a new campaign concept the Twin hasn't been exposed to, a style guide it should respond within, or internal data you want it reasoning against directly. Adding it here keeps that material traceable the same way everything else is: a response drawing on an uploaded asset should still be something you can verify, not a claim floating free of its source.
Keeping the Foundation Honest
The value of a Twin's grounding depends on knowing exactly what's in it. Reviewing Assets before a session, alongside the profile card described in E1. Conversing with AI Twins, is the simplest way to make sure a conversation stays anchored to real material, whether that material came from a Behavior, Intent, or Mentions report, or from a file your own team just uploaded.