IBM Watson “super-computer” – possibilities for procurement

At the recent IBM customer workshop we heard a lot about Watson, the super-computer. So what’s it all mean to procurement? Well, technically, “Watson is a question answering (QA) computing system that IBM built to apply advanced natural language processing, information retrieval, knowledge representation, automated reasoning, and machine learning technologies to the field of open domain question answering”.

So it handles vast amounts of information, and can then answer questions posed in natural language rather than programmed machine code that require it to apply the sort of thought processes humans would use – looking for links between items in the memory, testing hypotheses, identifying priorities and so on. All the stuff our brains do automatically.

The advantage over our brains is that Watson can hold every detail of millions of documents for example in its “head’ at once. So one recent application is in cancer research. It can’t “cure cancer”, but Watson’s cognitive computing ability will help go through all available clinical trials (over 100,000 around the world) and ensure that patients are accurately and consistently matched with promising and appropriate trial options.

There have been other more light-hearted applications. Watson won the US TV quiz show Jeopardy, and there has also been some publicity around its ability to create new recipes for dishes and cocktails! But there are more serious possibilities too.

IBM are now busy commercializing Watson, with three divisions being set up. Watson Analytics is the one of most interest to procurement, and will focus on Big Data visualization and insights on the basis of natural language questions posed by business users.

So one use case mentioned at the recent event was in a very large and complex public sector organisation, where it can take procurement people years to understand and learn all the processes, regulations, terms and conditions and so on. Watson enables users to ask questions and get instant answers around any aspect of the work, meaning people can achieve a level of competence much more quickly, reducing risk and increasing efficiency.

Another example might be around advanced spend analytics. We might still want our traditional “spend cube” and all the numerical reports that we’re used to, but Watson may be able to combine that with many other data and information sources to answer more complex questions. Contract creation and management is another potential area of interest. So we might pose questions such as:

“Are there spend categories where we are vulnerable to external risk”?

“Are there internal spend patterns in our organisation that might suggest fraud or inappropriate spending”?

“What contract clauses should we use the solve these particular issues”?

It’s going to take a while, I suspect, for us to work out just how we can use Watson, and indeed to see if it really is a useful or essential business tool for procurement - or just a super-powerful Wikipedia / Google equivalent. IBM also will have to work out just how to incorporate its power into the Emptoris product suite. But it certainly looks interesting, and the firm is putting thousands of staff into this area, so it seems likely that this is going to lead somewhere for procurement. Watch this space!

Share on Procurious

Voices (3)

  1. Ian Heptinstall:

    Are that many of us in a position where all the low-hanging fruit has been harvested, and the basics as fully embedded into day-day practices, that we really need massive computer power to identify what to do next?…

  2. Paul@Provalido:

    Sounds like Prof. Lamming’s long-standing theory involving a box, a man, a dog and a plane may become reality!

  3. Dave Orr:

    Don’t mention the Met Office then! #

    A £30m IBM supercomputer has been crunching weather data with mixed results when looking forward 3 months:

    – The infamous washout bar-b-q summer

    – The warm & mild winter that was one of the coldest on record

    – The 2014 “drier than average” prediction for Feb-April when Feb was the wettest on record and large parts of Somerset disappeared under flood waters for weeks & weeks!

    The old adage GiGo applies where without good base data and an accurate model of the real world, you will get fast computational answers – but the wrong ones.

    Good news on the Met Office front (excuse weathery pun) – they are buying a £100m supercomputer; getting better base data and the weather simulation model has been revamped:

    Now we can all now look forward to a new era of accurate monthly & quarterly forecasts and better micro forecasts across the UK to predict local flooding etc……

    An open EU tender has been instigated.

    This is one of those strategic technical areas where I would like to see (Airbus-style) a European solution, rather than simply buying in US technology.

    Fortunately, the new EU tendering rules allow those geo-political and socio-economic factors to be taken into account.

Discuss this:

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.