I think the other thing that it's very good at is giving us the surety around the output. So, once we've done our rate changes, we can very quickly promote them to market. And that gives us the ability to react very quickly to market changes, which in today's motor market, is absolutely essential.
I think one of the things that's often underplayed when you're talking about the development of machine learning within Radar is the impact that it has on the people who use it. There's lots of publicity out there about the sophistication you've built in, but it's been really good as we have recruited in our team that we're able to give them the right tools to do the job.
So when it comes to Python and looking at machine learning techniques, those guys are able to deliver the best by using the system that actually has that capability inbuilt. And certainly, over the last couple of years, we've seen that ability ramp up as you've worked on some of the more machine learning technologies and embedding those on Radar.
WTW have been fantastic. We made the decision before we went live that we needed assistance in terms of building out that initial rate book. Not the cheapest in the market, but I think the adage of you get what you pay for has come through. We work very closely with the team that were with us and the nice thing is that team have stayed in touch as we've gone through the last five or six years.
So we have fairly regular engagement now, although we have moved a long way from the original rate book and we've brought new data factors in, and we've developed the rating sophistication a long way from where we started out. We still have that quarterly contact with the team, and we still have conversations about 'what ifs?'. What if we did that, what if we did something else. And it's been really nice to have that continuation of support.
I think the other thing with the WTW team is that they have stayed in touch with us, so they understand where we are in the journey, because making the decision to go with a player like WTW right back at the start, when we were effectively a cash strapped startup, was quite a big call. But retrospectively, it was absolutely the right one.