Am I the only one who wants to export my eval results to a spreadsheet?

Okay, I have to get this out there because I'm genuinely curious. I've been using Freeplay for a few months now to run evals on our product's AI features—think summarization accuracy, tone matching, that kind of thing. The platform is fantastic for designing the tests, running them at scale, and seeing the aggregate dashboards.

But here's my thing: once I have those results, I feel... stuck? The charts and graphs are great for a high-level view, but I have this deep-seated, almost obsessive need to get my hands on the raw(ish) results in a format I can *manipulate*. I want to slice the data in ways Freeplay's UI might not anticipate.

For example, last week I ran an eval on our new query classifier. The overall accuracy looked good in Freeplay, but I had a hunch that performance was tanking for a specific user cohort (our enterprise tier on legacy data plans). To really prove it, I wanted to:
- Cross-reference the individual test case results with our internal user segment list (exported from our data warehouse).
- Calculate the accuracy per segment myself, maybe even run a quick statistical significance test in a separate tool.
- Build a custom scatter plot of confidence score vs. correctness for a specific failure mode.

I couldn't do any of that without a manual, copy-paste nightmare from the results table in the UI. What I *really* wanted was a simple "Export to CSV" button that would give me a row per test case run, with columns for:
* The test case ID and input
* The expected and actual outputs
* All the evaluation scores (from each judge/judgment)
* The model/provider used
* The timestamp

Is this just me? Am I being overly old-school, clinging to spreadsheets? I love modern SaaS tools, but sometimes you just need to take your data for a walk outside the walled garden.

I know there's probably an API I could use to fetch this, and that's my next stop, but that feels like a workaround for a pretty fundamental need. For a product built for data-driven teams, shouldn't seamless data export be a first-class citizen? I want to believe I'm not alone in this.

What are you all doing? Are you living happily inside Freeplay's analytics, or are you also finding ways to pipe results out to BigQuery, Sheets, or a BI tool? If you are exporting, how are you doing it?

🔥