Hey everyone, new to the forum here! I've been trying to learn modern data pipelines (mostly Airflow and dbt with BigQuery) and I keep seeing ads for Ideogram's new 'data storytelling' feature. It's supposed to automatically generate narratives from your charts and datasets.
Honestly, I'm a bit overwhelmed by all the tools out there, and this one has me curious but also skeptical. My current workflow is: I write my SQL transformations, build some Looker Studio dashboards, and then I have to manually write up summaries for my stakeholders. It takes me forever and I'm not great at writing.
So my question is: has anyone actually used this feature with a real data pipeline output? Like, if I hook it up to a BigQuery table after a dbt run, does it produce something actually useful for a business report? Or is it just a gimmick that gives you vague, fluffy text?
I'm worried it might just add another complicated layer to manage. But if it can save me time on those narrative summaries, maybe it's worth looking into? Would love to hear from anyone who's given it a spin with actual engineering data.
Haven't used it with dbt/airflow specifically, but I tested a similar storytelling tool on a basic GA4 dataset last month. The narrative it gave me was... okay? It correctly said "sessions went up by X%" but totally missed a key reason why, which was obvious from the chart.
It can maybe get you a first draft to edit, which saves some time. But I wouldn't trust it to spot a complex trend or explain a sudden dip without human eyes on the data first.
Might be worth a trial just for the summaries, if your dashboards are pretty straightforward? Let us know if you try it!