dataspan.in : ABS (Asset Backed Securities) Analysis – Instant and Effective

dataspan.in is a unique product. What makes this unique is the fact that it is able to successfully parse 80 ABX deals in their pdf format, extract meaningful data from these files and then create comparative data across deals for several parameters, such as delinquency at a tranche level. It also allows comparison across previous months. All of this within minutes of the data being released into the public domain.

What makes this process very interesting is its scalability. Currently it takes analysts around 800 minutes to extract data from 80 files at the rate of 10 minutes per file. And since its a manual process, several errors are typically committed. dataspan.in, uses an intelligent and automated pdf parser that does the process at the rate of less than 10 secs / file. Now if you had to do 1000 deals, manually it would take you 10,000 minutes, which is 166.67 hrs. Even if 3 people worked together, it would take 7 days. And then they'd need to redo the process just to check for errors.

Using the dataspan parser, this process would take less than 3 hours, using 1 machine. Use 3 and you'll have an error free output in under an hour.

And the best part of this whole process is that it's methodology is inspired by God's methodology. When we set out to solve the problem, we were faced with a major issue. The pdf parser was required to extract data from a set of rows and columns. Problem is that the pdf format does not store data as rows and columns. That's when we asked ourselves a simple question. Maybe the pdf files does not know, but U do. When I look at it, I can see rows and columns. So why not use the logic that my eyes use, to create virtual rows and columns. And thus was born the "Digitizing Transmuter", the intelligent pdf parsing engine.

You can see the input pdf files and output ABS deals data at www.dataspan.in

This product seems to have applications across verticals. And in case you think, it can help in your field of work drop us a line, and we'll be happy to discuss the possibilities.