I’ve made no secret of my scepticism regarding AI reliability. But this week, SuperAgent might have officially changed my mind. 🤯
I’ve been back in the "vibe coding" lane within Airtable; the Omni upgrades of late are producing some stunning interface results. Now it was time to test out SuperAgent. I fed it a dataset I know inside out: the Product Ops Confidential Product Operations 2025 Salary Survey. I asked for a comprehensive report focusing on location, role grade, experience, and company size. The prompt was not much more than that (I did ask it to refine 3 items after the first run: Add in cautions on segments with low responses, make a GBP version as well as the USD version, and correct where one of the role types stood in the hierarchy).
The results? I’m "SuperImpressed." 😅
🫡 Depth & Logic: It handled nuanced prompts, like flagging where low response counts might skew conclusions.
📈 Statistical Rigour: It surfaced insights I hadn't even spotted in my own dashboards.
⏩ Speed: A professional-grade report that would have taken me hours (and skills I don't have) was finished in 5 minutes.
Here is just one of a dozen charts it included in the report for me.
I'm taking some time to review the outcomes before considering whether to publish the full report... watch this space!
