IT services giant Accenture has tens of thousands of consultants around the globe. Many of them may be working on similar projects at any given time, so sharing knowledge across the organisation has considerable value.

To improve the cross-pollenation of knowledge and innovation between projects, Accenture Labs developed a knowledge- management system based on a taxonomy to categorise projects. Crucially, the system allows projects with fundamental similarities to be linked even if they are not superficially related.

Building the systems has allowed Accenture to decommission scores of internal databases, cutting annual IT costs by over $1 million.

It was always developed with one eye on commercial exploitation, says Michael Redding, managing director of Accenture’s Technology Labs. “We had a basic idea that if we could develop a system that improved our ability as an organisation to tap the expertise of our staff, it might also be of interest to our customers.”

But Accenture Labs did not anticipate the uses it has found outside the organisation.

The system’s ability to draw non-obvious links between items in a taxonomy has proven especially useful in the pharmaceuticals industry. “Our teams in the pharmaceutical sector realised that this ability to make hitherto unseen connections between things had real potential,” says Redding.

The first demonstration of this potential was in identifying similarities between gene sequences in patients suffering from Alzheimer’s disease and those with alcohol problems. It has since been used to spot a relationship between heart failure and leukaemia. Another discovered a link between Lou Gehrig’s disease and genital diseases.

“There is a real chance that drugs used to treat one condition could have beneficial uses for conditions that they were never previously considered for,” explains Redding.