In an earlier post I reflected on how far RELEX has come as a company. It made me realize that our values are much the same now as they’ve always been. Our customers’ success and our own are indivisibly linked. It’s all about a job well done.
Key to that has been our technology. From the outset we have strived to combine our in-depth knowledge of supply chain with our understanding of software systems. And while our work philosophy has been pretty constant, the only constant with our technology has been evolution.
If we’d started RELEX in a large European country, say Germany or the UK, the technology might have evolved very differently. In Germany, for instance, there are scores of mid-sized companies in almost every sector. But in Finland there are rarely more than 2 or 3 players in any given area of business. To have developed a specialist solution that simply served one sector would have been to have raced up a blind alley. The need to generate cash flow by catering to a wide range of, often very different, companies, led pretty directly to what we have today. When we told Marko that we needed to accommodate a multitude of sectors he was pretty blunt: “I can’t be bothered to program separate solutions for all those,” he said. “I’ll build you a platform that you can adapt so you can solve each customer’s challenges yourselves.”
From the outset we have strived to combine our in-depth knowledge of supply chain with our understanding of software systems.
Our first columnar-in-memory element was already in place by 2007. This columnar database extension provided us with a free-dimension, in-memory UI. It was ground-breaking and gave us unprecedented flexibility to arrive at solutions with customers. However, it also had limitations. The UI data was in a separate database from the backend. This means that the data that one saw was not the data used in calculations if you didn’t run batch updates. The manual adjustments, such as (manual) forecast corrections, were also quite slow. Finally our biggest customers were putting the relational database under pressure. The index tables alone were often well over 1TB in size and so, typically, were forecast tables.
So, of course, we decided to develop our own database, one that could be optimized for the data structures and application types that dominate our business, and that could be integrated deeply with the solution to facilitate further performance improvements. At first Marko had been hesitant to push forward with this idea: “I did not want to say out loud that I thought we should build our own database. People would have probably thought that I’d finally and completely lost my senses.” That was understandable because not only was it a big challenge, but we also decided to rewrite the UI (which meant redoing over 80% of its code) at the same time and move to a slightly thicker web-client for faster performance.
We decided to develop our own database, one that could be optimized for the data structures and application types that dominate our business.
‘The New Version’ took longer than we’d estimated and, for an organization used to getting important developments finished in days or weeks, spending two and a half years to complete one project was a cultural challenge. It also came as we were starting to internationalize the business and taking on some of the largest retail accounts in Finland. We did well to come through that period with the business in good shape and with happy customers. The Solutions Team did an amazing job keeping the ‘Old Version’s’ heart beating while on its deathbed; think George Clooney in ER except with SCM software. Michael had a particularly tough task leading our Dev operations. Every other word he had to use was ‘no’. If he wasn’t saying ‘no’ he was probably off duty.
If you weren’t there you’ll probably have a hard time imagining how wonderful it felt to behold the beautiful piece of software engineering that was the new version. It was equally wonderful getting to grips with all the ways it could be used to help our customers. The graphic UI of the Optimiser was revolutionary (and revelatory) and the data update performance was a million times faster. It was bliss.
Looking back we made the right choices and were rightly proud of the service and results RELEX delivered for its customers. I like what we did and how we did it. We were the first vendors in Europe to crack the difficult challenge of fresh product replenishment. We saw that the combination of flexible technology, talented people, and innovative team spirit would stand us in good stead in taking on, pretty much, any challenge.
And that’s why, late in 2014, we decided to see if we could take RELEX global. It’s all very well being a big noise in Europe, but we’d like to see our approach, our people and our technology go toe to toe with the best the world has to offer. I think we’ll give a good account of ourselves.