We asked Dan Sommers, founder of Warwick Analytics, to share some wisdom on starting a company with members of the Microsoft BizSpark program.
I love challenges and to do better with each new business I’m involved in. When I felt the calling to launch Warwick Analytics, I knew we had to go for it because the idea just got better every day and every time I did more market research. It’s where academia meets the real world, a rare union, and so incredible, really, that I didn’t believe it at first.
I met Professor Darek Ceglarek of the University of Warwick and his team at a time when I was searching for a business opportunity that could really make a difference and attract viable markets. As Chair of the Engineering and Physical Sciences Research Council, he was involved in a lot of leading edge research work in manufacturing.
The search process for a strong commercial opportunity could be termed a ‘frog-kissing exercise’ – there are so many people with good ideas but it takes time to find ones that have commercial potential and a personal fit. However, one piece of technology that Darek had been involved with was different from anything else I had seen. He and his team had been working on ground-breaking technology that automatically locates the root causes of manufacturing faults and recommends the most beneficial actions without the need for hypotheses. It can work with incomplete and dirty data and even resolve no-fault found problems.
In the most basic terms it allows engineers and manufacturing personnel to do their jobs faster, more efficiently and with enhanced product quality. As I spoke to potential customers, I quickly realized there was a significant commercial opportunity to explore. We could potentially save the manufacturing industry billions of dollars each year by reducing their Cost of Poor Quality (COPQ). For example, in speaking to the Head of Quality at a well-known automotive manufacturer I realized that his team was dealing with thousands of different issues at any one time and using guesswork and statistical tools to try to figure out the causes. I was hooked.
We’ve now patented the technology as RCASE, Root Cause Analysis Solver Engine and I’m proud to say its winning awards around the world because of the unique technology involved and the way it finds manufacturing defects.
Of course, more than academic excellence, we needed a strong commercial management team so we brought in Jason Noble as CTO to manage technical strategy and Mark Hardy as Chairman to help with strategy. We complement each other nicely with our individual strengths – it’s critical for a startup to have a management team that works well together.
We chose to work with Microsoft using a Windows Communication Foundation (WCF)-based cloud solution that can be deployed to Azure. Our algorithms split into different stages and each stage can be parallelized. Plus, we also need to run a lot of scenarios where we tweak input data and parameters. It’s critical that we have a scalable compute framework and platform; Microsoft’s tools and services give us exactly what we need.
Without Windows Azure, frankly, we would never have fully validated our business model. Building in the cloud should be a no brainer but there are still negative perceptions about security that need to be overcome. Windows Azure and our partnership with a worldwide technology leader like Microsoft through BizSpark go a long way toward dispelling those negative reactions because of the reputation Microsoft has. People trust Microsoft to get the job done.
We may be growing fast and deploying our technology on a global basis but like most startups, we can’t do all this on our own. Some of our large corporate partnerships – most notably Microsoft and SAP HANA – have been almost planned serendipity.
I think that if a startup can find a company that will take you under its wing and help you move forward, it’s something you should take advantage of. Even if you’re told ‘no’ don’t let it stop you, just keep making calls. If your idea is strong enough, you’ll get the endorsement you need.