2 mins read

As part of the data engineering team, we look at data with a few different perspectives. One is through the lens of engineering and transformation, getting the data from A to B and putting it in the correct format, so that we can build products out of them. Another perspective is looking at the data holistically, to tell stories and use the data in ways we haven’t been able to before.

The really fun bit is when we are combining different data sources that are not traditionally used together. We are doing this to leverage new techniques like machine learning and artificial intelligence (AI), so that we can achieve more out of the building blocks.

AI and the methods behind it are not new to us. These methods have been around for 30 to 40 years but are largely inaccessible due to their immense computational requirements. It’s only in the last few years that computers, particularly cloud computing, has reached a scale that we can actually use AI in everyday applications.

The biggest impact is in areas such as image recognition, video detection, and human language processing. Another way to think about AI is: What we thought only humans could do, the machines can now also complete. This is allowing a new kind of automation.

AI-assisted automation could be applied to video feeds, media reports and twitter messages. It could also be applied to complex tasks that we expect only a human can perform. Using AI techniques, we can combine different data sets, look at patterns and inferences, and deploy many, many robots. We can even teach robots English to interpret messages and extract key words. We have developed AI-assisted automation to parse traffic incident reports, which has allowed us to achieve a higher quality of service, quicker response, and expansion to a larger scale that we couldn’t manage before.

Another exciting application is bridging the gap between multiple modes of transport. In our Insight product, we handle a lot of traffic, transport and mobility-related data. We want to be able to understand how people move around, what modes of transport they prefer, and the environment around them, such as how weather affects the choice of how to get to work.

We want to be able to recommend, in response to current events or incidents, whether you should drive or take public transport, or whether you should work from home today. We want to be able to say that today, you should go home a little bit early, because the weather is going to be bad and you might want to walk your dog.

We want to use data to make people’s life easy.