Then there's a simple interface to match up key fields to optimize the results. You can keep experiment until you get the results you need.
So the challenge is how do you make a software solution that will work with the less technically proficient user. In today's world, everyone is very comfortable in a browser. And less comfortable with Excel. So the browser as a software platform is a good solution.
Initial work on Crush deployed many features. First there was a comparison engine. Second, there was a way to line up data on key columns. Third, summation of balances were added.
But then it was clear that some columns needed to be added, like a column that would subtract CREDITS from DEBITS. Obviously, this formula could be added in Excel. But the goal has always been to AVOID work in Excel and ideally work directly from a CSV. If Crush depended on the user making changes in Excel, the value of the system would be reduced.
Then it was clear, more was needed. It was essential to filter out certain rows. For example, a CSV from a bank could include THREE months of data, but to do a monthly bank reconciliation, only ONE month could be used. Again, those other two months could be deleted in Excel, but again, that is not the ideal.
Recapping, there are lots of pieces to do a reconciliation. A difference engine. A key column engine. A summation engine. A way to add columns. A way to filter rows.
That's a lot to handle. So the next step was to build a recipe system that would tell Crush what steps to take -- step -- by -- step. That really worked very well, but it is essentially a feature that would require yet another level of technical know-how. But for technically oriented users, it is great. It essentially automates all the steps needed to Crush errors.
So the missing piece -- that is finally done -- is the SETUP WIZARD. Now the NON-TECHNICAL person is guided through the process step by step to find errors.
There's a lot of artificial intelligence. First, it eliminates columns that are essentially empty. Users tend to put in empty columns between real columns. That makes the system a little messy.
Then it looks for columns that are named Debit, Credit and dozens of other standard column names. It then figures out what columns you may need to add. It asks you to do that. It then asks which rows you need to filter. RIght now, it only filters by month or the CLR column. The CLR column is produced by QuickBooks. If there's an asterisk in the column, then it should be included in a reconciliation.
Then it examines columns that have the same name and it assumes they are key columns. But it will also pair Amt with Amount.
Then it color codes each column. Columns with number values are green and have a green box. Columns that are added by the system have a blue box. Crush makes suggestions as to how to organize the fields. But users can now easily modify the suggestion by dragging fields to the right box. A Key field, like Customer, could be dragged to the Key box. An unimportant field, like zip code, could be dragged and dropped into the Ignore section. To drag all the fields, hold down the ALT key.
So to recap. There's the artificial intelligence that added columns and filtered rows There's the intelligence that recommends how to organize the fields. There's the easy user interface with color coding and drag and dropping.
But the last new essential feature is that there is a TEST button. So you can move a field and then TEST the results. Then there's a popup that appears with the results. Then try another combination of fields. And then do another test. Compare the results. Every popup has a button that will restore the combination of fields to the one that generated those results.
So users can keep experimenting. Yet every experiment keeps track of all the steps. And displays the results. So you can restore your settings at any time to the same settings that produced those results.
It is now an integrated solution with all of the power required to solve a hard problem and the ease of use that makes it accessible to all.