Yesterday I looked at a recent project by a collaboration of researchers that explored whether court judgements could be predicted by automated systems.

It’s part of a growing march into professional services by automated algorithms that sees no sign of abating.  For instance, recently IBM announced that Watson will be deployed to help organizations manage the maze of financial regulations that they have to abide by.

The company recently purchased Promontory, a financial consultancy that has been working with the SEC, World Bank and others in the financial regulation field.  Their core model is to help clients manage the dense fog of regulations via their team of experts.

Codifying expertise

Suffice to say, Watson is an automated solution, so the partnership with IBM will see an attempt to codify the expertise of the Promontory staff into the core Watson decision making engine.

Whilst the task is a considerable one, so is the bounty should they get it right, with expenditure on regulation and compliance believed to be around $270 billion per year.  Indeed, around $20 billion of this is just on understanding the rules in the first place.

A pilot program has already been run with a number of banks, with Watson hunting down a number of potentially illicit schemes hidden in the data it was fed, which ranged from trading patterns to social media.

The researchers next hope to add support for rules to the system, with users able to cross-examine Watson to gain insights into areas ranging from jurisdictions to products.

During this initial phase, the guidance offered by Watson is itself vetted to provide it with a degree of feedback so that it itself can learn and improve.  With speeches and other content types soon to be assimilated into the dataset, Watson will have rich pickings to mine in it’s hunt for insights.

The eventual aim is to provide the best form of real-time support to companies looking to stay on top of the seemingly ever changing web of regulations.  This will be especially crucial for smaller companies that lack the resources to hire the kind of expert teams available in larger companies.

It provides another example of how AI is beginning to influence professional services as well as more manual work, and it will be fascinating to see just how successful Watson becomes in this crucial field.