Vodafone drafts in Hazy to trial ‘synthetic data for telecoms’

UK operator Vodafone has been tinkering away with a firm called Hazy to test the use of ‘synthetic data’ in its operations.

Vodafone brought in Hazy to run a proof of concept project looking at how synthetic data might be used for training and testing machine learning models, and the operator is now mulling how it could adopt said synthetic data capabilities into its operations.

What is synthetic data, you might ask? Broadly the term means data points that are not drawn from the real world but generated by an algorithm in order to feed something like a machine learning programme, perhaps for modelling a particular scenario for which the actual data is not on hand.

The announcement says it allows companies to digitally generate the data that they need, which can speed up Big Data projects, and in this case synthetic data was leveraged for training and testing machine learning models for customer value management. Currently its machine learning models are trained using large amounts of customer data, which can be a slow process apparently.

Vodafone’s R&D team used both internal metrics and the Hazy software built-in metrics to evaluate the generated synthetic data, concluding that ‘the synthetic dataset preserved the performance of the machine learning model, saving the team time, cost and in some cases even increasing the model performance.’

“We are really excited about the potential of Hazy Synthetic Data,” said  Luke Ibbetson, Head of R&D at Vodafone. “Our team was impressed with the quality, utility and usability of the generated datasets, as well as the software’s ease of deployment. We are keen to explore the power of a synthetic data platform for Vodafone teams to access useful synthetic data. This software could be a powerful, cross-function tool, saving time, increasing productivity of data science teams, accelerating project deliveries, and enabling us to operate more efficiently.”

Harry Keen, CEO at Hazy added: “We are delighted with the success of this Proof of Concept and excited about the potential of synthetic data for Vodafone. Thanks to the close working relationship developed with the Vodafone team and a great choice of data set for customer value management, we were able to rapidly develop a synthetic version that met the success criteria for privacy, speed and fidelity. We are looking forward to building on this success to operationalise the value of synthetic data across the group.”

The long and short of it seems to be that generating synthetic data to run an algorithm took one minute, as opposed to ‘potentially waiting several weeks to get access to production data or create a test dataset manually.’ The obvious thing to point out would be presumably there is some sort of a downside if the data is made up, in that it doesn’t exactly map onto or reflect the reality of any given situation.

Regardless, there are probably all sorts of areas where such tech could be applied, and Vodafone appears to have concluded ‘customer value management’ might be one.

Get the latest news straight to your inbox. Register for the Telecoms.com newsletter here.