Working for Vodafone on behalf of my client, Accenture, as part of the Performance and Chaos Engineering team; our mission was to develop an automated performance testing framework for our applications to support the many daily production releases.

We won "most innovative project" for training an ml model to detect service degradation and identify the peak load per target, which could then be used to automatically generate scaling configurations for the applications.

I'm particularly proud of this one as this project was my baby; I spent several weeks running tests around the clock. I even had to write an interface to be able to visualise and efficiently mark up hundreds of sample sets of data manually. We ran a hackathon in partnership with and, fortunately, I had a couple of really bright engineers out of university with backgrounds in ml, that were excited enough to volunteer to join my project. Within 3 days we had a model that was 95% accurate!

Not only were we recognised by in the hackathon as the winning project, but was also recognised here by the DevOps Industry Awards.

I was also delighted to be invited along, even though I had moved to a new contract 8 months prior (to help clear all the booze off the tables .)

The above ml project was recognised for "best use of AI in the cloud" by The Channel Cloud Excellence Awards!

While Working for Vodafone on behalf of my client Accenture, our automated performance testing platform was recognised as "cloud development platform of the year" by the The Channel Cloud Excellence Awards!