technically metadata is just "data about the data". So its all just framing. One ETL pipelines data could be another's metadata, and vice versa. For example, let say her "threat intel" is just twitter usernames. Then the metadata would be the "second tier" of data. For example, lets say she threw twitter usernames into an excel sheet. The data is anything regarding the first order dataset. So anything involving the twitter users, their username, the tweet, the replies, at mentions etc.
It isn't clear what she means by metadata in this context. The metadata would be information about those excel cells and the pipeline that built it. The service account that scraped the data, the time. The RED metrics (rate, errors,duration). More commonly she'd be referring to the ETL paradigm and the schema or structure of the data. So metadata here could be something like the row headers, excel sheets, vlookups, and pivot tables needed. In a more production setup it would be the db schema etc.
This isn't a fucking arch spec though so. I can't imagine why she'd be saying "metadata" in this context. I certainly could be wrong about this I'm not the arbitrator of tech jargon, I've just never seen a professional reference metadata in that sense. She's welcome to correct me, but somehow I don't think she will.
She probably mistakenly thinks metadata just means "more granular data", instead of second order abstraction of the data itself. Because she doesn't have actual use case knowledge and just uses the jargon colloquially. Shes fat and dumb.
She probably thinks data = the twitter user name. And metadata is information about the user or the tweet. But that's all just data itself.
Say data again, stupid.