Along with our partner AWS, we participated in last week’s KPMG Financial Services TechOut event at Canary Wharf. It was a great opportunity to meet KPMG customers and staff in an informal networking environment. The purpose of the event was to showcase new and disruptive tech ideas to KMPG financial services customers and staff alike, and so there were stands from various internal KPMG teams (e.g. blockchain) as well as a few select external organisations like us/AWS.

KPMG TechOut event

One of my victims…

As you can see from the photo, we gave some live demos.  On the right was an example of stream-based analytics running on our RAMP platform, performing “on the fly” machine learning predictions (a regression model) with multi-dimensional anomaly detection.  On the left, we were showcasing facial analysis underpinned by AWS Rekognition, which if you zoom in you can see is determining in real time a number of attributes such as gender, sentiment, age etc.


Interestingly, the audience typically showed the most interest in one of the two demos, i.e. they could see the applicability of image recognition or business predictions from structured data to their life/job/business.  AI is a broad church spanning from classic machine learning through to deep learning (and the expansion of this church is far from over yet – e.g. see this at Capsule Networks from Mr Hinton and co).  So it’s good to see people naturally gravitate towards whichever demo that they can see can make an impact in what they want to do – after all is said and done,  it’s all about application of these technologies to a real use case.

On a personal note, I enjoyed rubbing shoulders with the KPMG people.  It’s a different crowd to the AI/ML bubble and I learned lots about what an auditor really does and risk management in the banking sector (for example).  Like most subjects, once you get under the covers, it’s actually quite interesting! 🙂