Comment by alexliu518

Very critical question. Excel is indeed more complex. Before analysis, Capalyze first preprocesses the data, which is crucial. We have designed a set of preprocessing algorithms that essentially focus on how to better identify the data suitable for analysis in Excel and clean and repair it. This process also leverages LLMs, as we found that LLMs perform quite well in recognizing table structures.