Can you tell us about your business?
We are a software company that makes it easy for banks to build Artificial Intelligence simulation. Our platform, Providence, allows banks to build models and use advanced algorithms to understand emergent behaviour in complex systems like the US bond and UK housing market. In terms of the UK housing market, Providence keeps track of all 27 million households. In the past, this work was done by averages. That seemed to make a lot of sense until the world economy blew up in 2008. Providence builds simulations to understand how different policies can have an impact.
One example is the buy-to-let market. A lot of people want to buy homes and rent them out as an investment but this can be very speculative. Prices can rise to a point where they become unsustainable. If the market pops, asset values fall, rental incomes disappear and people who were relying on that income to take them into old age go underwater, unable to sell for more than they owe. Providence really allows you to understand subtle and complex issues. Changing the loan-to-value or loan-to-income ratio by a central bank has an effect. In order for the AI to beat humans, it has to play the game millions and millions of times. But in that process of playing, it learns.
In the financial world, there are literally trillions of dollars at stake if the wrong decisions are made. It’s the “black swan” event – an occurrence that comes as a surprise – that simulation is really good at. It allows you to keep track of all the moving parts in a market. Our tool is used by the banks to help them with their strategic decisions both on the risk and the money making side.
Where did the business idea come from?
I started working for the US Department of Defence in Afghanistan and Iraq in 2001. I focused on simulating insurgencies and counter-insurgencies. I moved on to building simulations for gas, oil and insurance companies and worked for the Bank of England for five years. My drive has come from a disappointment at how our leaders make decisions.
What is the potential of Providence?
At Simudyne, we want to simulate the entire world. The financial sector is just one vertical. We’re going after oil, gas and transportation. We’re building a Star Trek computer. You don’t want something that works one time in 1,000 – you want a strategy to ensure it works 800 out of 1,000 times. And even where it gets it wrong, it should alert you quickly so you can change direction based on new data. Simulation has a proven track record, allowing you to put in place mitigating strategies to prevent the downside. You might lose money but you’re not going to go out of business.
What’s next for the company?
We’ve signed our deal with Barclays and we need to gain good commercial traction. We want to raise significantly more capital, add staff and expand. My dream is that where now people search Google for information, in the next five years they begin to ask Providence for answers. We want to get in touch with the consumer. If somebody is buying in a certain area, they could ask Providence if the risk is too high – and it might suggest a different area. We’re not building a crystal ball. This is not magic. It’s just applying mathematical modelling to these AIs.
What did you get out of your time with the Barclays Accelerator programme?
The Barclays Accelerator programme was utterly transformational for us as a company.
We’ve raised significant capital. I’ve never worked so hard in my life. Even when I was in Iraq, it was not as intense. No one was shooting at me but on the Accelerator nobody would let you say something without challenging it.
I really appreciated Barclays being so open. Normally getting that kind of feedback from a bank is incredibly difficult. In the three months we were in the programme, we got more than 100 meetings. You are prepared to have your ideas challenged. There were many times when our dreams were crushed. But you had to go back and say: “What’s next?” That sort of feedback for a start-up is precious.
As someone born in Texas, I feel perfectly content making fun of cowboy decision-making. Cowboy doesn’t work in the 21st century. You’ve got to be more measured. With what is happening around the world, we need Providence more than ever.