Thanks for your review. I would love to know what specifically you were looking for. We develop our platform responsively and put a great deal of effort in adding new features for our customers use cases.

In general, getting an advanced analysis depends on the size of the dataset you drop. We provide advanced analysis in many ways inclusive an easy explanations on what they mean in the context of your data and use case. Here are some examples and a small portion of that ChartPixel does:
- Regressions analysis
- Trend analysis including polynomial fits
- Forecast through advanced models such as ARIMA, exponential smoothing and half a dozen more autoselected
- Significance t-tests and z-tests for probabilies to know if your results are sound
- Chi squared tests for finding oddities in your charts to understand the drivers in your data
- Automated detection of ordinal categories through semantical analysys
- Trends in categorical data
- Outlier detection and checks for extremes through more statistical tests such as Cohen's d
- Feature importance of specific categories including Shap and direction of contribution
- Topic and sentiment analysis from comments
- Panel data analysis including reformatting of your dataset to create panel charts
- Kernel density plots and Ridge plots including an analysis of those
- Various machine learning models that help you to understand what drives your results

Note that for small summarised table uploads or uploads with few columns and rows, analytic insights are more limited as there is often no statistical base to generate more advanced insights from it without getting "speculative". We are not a LLM wrapper and do not drop your data into these to wait for an answer to output. We have build ChartPixel to provide genuine insights through our domain expertise and models beyond that and are always happy to learn and implement more.

Thank you again for your feedback and if you wish to actively help us developing or you want to requests a feature, please contact us at info@chartpixel.com.