Daniel Hulme is the founder and chief executive of a young British company specialising in optimisation algorithms and services. It’s akin to a maths think-tank. Satalia’s staff of academics and artificial intelligence or AI experts harness a library of algorithms to help companies solve the world’s most difficult problems. It was recognised this year on the elite Gartner Cool Vendors in Data Science list, the only UK company chosen.
It helped a major UK telco embark on a big upgrade programme. Copper lines would be replaced with fibre. With thousands of miles of trunking to consider, the telco needed to calculate the best possible routes. Even tiny improvements to plan the fibre routes would mean huge cash savings. The problem? The number of possible routes to consider quickly spirals out of control. There are trillions upon trillions of conceivable sequences. It’s insanely hard to compute.
In fact, the problem will be familiar to anyone with an engineering background. It’s a twist on the travelling salesman problem taught to every computer science and advanced maths student.
Satalia ran the telco challenge through its algorithms library they call the Solve Engine, which uses machine-learning to select the best algorithms to solve optimisation problems. The Solve Engine found a solution 1,000 times faster and 20 per cent better than the existing approach.
We can tackle any type of optimisation problem, which usually involves some combination of planning, scheduling and routing
The example is far from a one-off. Satalia is at the forefront of bringing algorithms to the masses. It has a growing portfolio of success stories. In each case Satalia uses algorithms to solve an optimisation problem too complex for the client to handle.
“We can tackle any type of optimisation problem, which usually involves some combination of planning, scheduling and routing,” says Satalia founder Daniel Hulme. “We have a library of thousands of algorithms and deep expertise in optimisation, which means we can help clients with almost any complex task. We deal with a wide variety of requests, including workforce scheduling, planning mobile phone masts, identifying bugs in software and creating the best vehicle delivery routes.”
Even small problems can be blitzed by the Satalia team. A university asked Satalia to create a timetable to fit student preferences and lecturer availability. The process created the most efficient timetable possible, while cutting the task time by 60 per cent.
Dr Hulme is evangelical about the potential for companies of all sizes to embrace algorithms to solve everyday problems. Back in 2007 he developed the idea of optimisation-as-a-service while doing his doctoral thesis in AI at University College London. He has since become a global figurehead of the algorithm revolution, speaking at TEDx talks and industry events on how algorithms can be harnessed.
His mission is to bring algorithms to the masses and help business leaders realise how critical optimisation is to the success of their company. “We’ve all heard of big data and advanced analytics,” he says. “The basic level is descriptive analytics, which is just the visualisation of data. Lots of firms do that. Then there’s predictive analytics, which is about making forecasts. A small number do that.
“We are offering prescriptive analytics. This is a new field in which algorithms recommend specific actions. With prescriptive analytics you now know what to do with the insights extracted from data.”
One of Satalia’s biggest successes is the optimisation of the delivery routes for a leading international retailer. The client understood the potential cost-savings of using algorithms to compute the best journeys for delivery vehicles.
Dr Hulme says: “Suppose you have to deliver packages around in a city. If you have 24 packages to deliver there would be 24x23x22x21 and so on (620,448,401,733,239,439,360,000) possible routes. Even if a supercomputer could calculate a million routes every second it would still take 20 billion years – longer than the age of the universe – to determine the best one. Add another delivery and it’s 25x20 billion years (500 billion years). This is why algorithms are so important in the rapidly growing complex world of business.”
For large corporates, we build integrated end-to-end AI solutions; for small companies, we take away the need to have expensive data-science and optimisation teams
So can any company approach Satalia with a problem to solve? “Absolutely,” says Dr Hulme. “We’ll take a look at anything and we’re motivated to solve the biggest problems. For large corporates, we build integrated end-to-end AI solutions; for small companies, we take away the need to have expensive data-science and optimisation teams.”
Satalia has clients from the United States to Japan, in multiple sectors. “We’ll soon be launching a public version of the Solve Engine,” Dr Hulme reveals. “Customers will buy solve time by the hour and every time they submit a problem, it will use up some time, similar to mobile phone pay-as-you-go credits.” The pricing model is flexible. Jobs can also be priced per problem, through an enterprise-level contract or with a permanent on-site presence. A freemium model is also available.
Awareness of algorithms is soaring. A lot of companies look at examples, such as Amazon’s product recommendation service or an airline’s dynamic pricing engine, and wonder whether it is plausible to create something similar in their own enterprise. Satalia makes it possible.
“Too many companies are either not using algorithms or using vanilla in-house versions that really aren’t fit for purpose,” says Dr Hulme. “Our Solve Engine and expert services means they can completely future-proof their optimisation.”
Satalia’s algorithm library is regularly updated by the academic community, using the latest innovations in optimisation. This allows companies to make the best decisions from data and focus on their core business.
Dr Hulme is regularly invited to events and conferences to speak about the role of artificial intelligence in tomorrow’s world. “As a company we don’t do cold calling,” he says. “We raise awareness about the huge impact of optimisation and algorithms, and how to solve incredibly hard problems. Then when companies realise what we can do for them, we become partners.”
To find out more visit Satalia.com or e-mail Daniel Hulme at daniel@satalia.com
Top image - Dr Daniel Hulme, chief executive: “I love talking about what is, and what is not, artificial intelligence”
