The agency has a vision for solving this problem, and it requires an integrated government transaction clearinghouse. If all financial transactions were to go through a single clearinghouse, and if that clearinghouse understood the details of all the transactions, then it would be possible to figure out where the money is actually going. It would be possible to optimize expenditures, so that more – possibly lots more – could be done with less.
This organization has been working toward realizing this vision for many years, but the problem is so big that it can’t be solved in one step, or even in one giant leap. They have already set up a number of transaction clearinghouses. Grossly simplified, these clearinghouses pass transactions back and forth between different agencies to get things purchased and to make sure those purchases are actually delivered.
But these clearinghouses have been designed with a “come-as-you-are party” mentality. In a normal commercial organization, a central governing body will often dictate the interfaces between systems, even those that cross organizational boundaries. If the target were a toga party, then everyone would have to wear togas, like it or not. However, in this organization, each department has a great deal of autonomy, with different systems and often very different security requirements. It is neither possible to reach into the different systems to get the data one wants, nor to dictate how the data will come dressed for your system.
For a transaction clearinghouse, the result of this come-as-you-are party mentality is that everyone comes to the party with a different set of incompatible interfaces and data, and it is the clearinghouse’s responsibility to make it all mesh together. One of the largest transaction clearinghouses in this entity has been growing in this manner for years. Unfortunately, it has reached the limits of the well-known EAI (enterprise application integration) product it has been using.
The core issue with the EAI technology is that it has no sense of “metadata”. The interface rules are specified by analysts on paper, and then coders turn the specs into programs that are plugged into the EAI framework. The EAI framework has no idea what is going on in the rules. The rules are coded in third generation programming languages and are therefore opaque to all but the original coder (assuming the original coder can even remember what he did). There is virtually no opportunity for reuse, and it is almost impossible to impose standards. This was all right at the beginning of the effort, but as the system has grown larger, it has effectively collapsed under its own weight.
Luckily, the systems integrator (SI) responsible for this transaction clearinghouse saw the collapse coming way before it happened and initiated a search for a new technology – one that would handle the transaction volume and complexity, that would coexist with the old system while it was being phased out, and that would not just capture metadata but also accept metadata as a specification mechanism. After an exhaustive search, the SI realized that Ab Initio’s Co>Operating System was the answer.
With the shift to the Co>Operating System, the SI has had to rediscover many of the existing rules, because none of the people who originally implemented them were around. Further, many shortcuts had been taken in the original system because there had not been a standard for the specification of a transaction. In many cases, the original system just passed a transaction along without having any idea of what was in it. That was clearly unacceptable and is no longer happening in the new system. Each interface has been redesigned with new standards in mind. Ab Initio also helped by providing automated translation capabilities from the old rules technology to the new.
After much previous investment and no success, this agency is finally getting the results it had always wanted. It is beginning to be able to analyze transactions to associate total cost of ownership with specific capital items. It is beginning to be able to discern expenditures with different activities (R&D, maintenance, operations…). It is starting to be able to answer prosaic questions like “How much does it cost to build a XXX?”
And when you’re dealing with financials on a monumental scale, that’s a very important question to have answered.