AI-powered text search

It is not just repetitive data input that takes up a lot of an Analyst’s time, but also the data search itself. Trying to find the right data point in hundreds of pages is sometimes akin to finding a needle in a haystack.

Exerica makes searching easier by filtering raw text in two key ways

  • Filtering results that do not refer to any numerical data sets
  • Taking account of the structure and hierarchy of numerical data (search for “tax included in operating expenses” or “capex as part of investing activities”)

To get a sense of how this works, think of a simple phrase like “capital expenditure” which can appear in different contexts dozens of times in an annual report. Usually a text search would be of minimal help, and the analyst needs to find the right note or chapter, then look for to the right disclosure, before locating the specific data point they were after.

An ordinary data search for most analysts can be far more complex, and they almost always involve the hierarchy of the data being presented, and the structure of the document.  For example, an analyst may be looking specifically for taxes reported as part of company operating expenses (a.k.a ‘taxes other than income tax’) or for CapEx disclosed specifically in the cash flow statement. Such search requests would normally take minutes rather than seconds, with analysts having to go through the entire document and look through various disclosures before getting to the right place, and the right piece of data.

Exerica does this for the analyst. It restores the original document and data structure, and processes complex analytical requests in exactly the same way analysts seek data.



Maxim Miller
March 4, 2020