Bonjour! Who is GenLots and what are you doing?
Arnaud: “We are a company from Lausanne, Switzerland and use machine learning to optimize supply chain processes. GenLots ́ main technology, the artificial supply chain brain, processes data for production companies that order raw materials in high quantities. Our algorithm improves purchasing plans mostly by decreasing order transactions and reducing inventory. With our software, industrial companies order the right amount at the right time and thus save up to 10% of total purchasing cost.”
How did the idea of GenLots come to life?
Simon: “During my master at ETH Zurich, a Swiss pharmaceutical company approached our chair of supply chain Prof. Wagner with a specific question: “Given our production plans: how much raw materials should we order when, to keep our inventory low and reduce the number of orders, without neglecting quantity discounts.” I was very fascinated with this question but realized soon, that neither academia nor business had a satisfying solution to it. So I decided to build one myself and developed the first version of the artificial supply chain brain for my master thesis. Eventually, I imported real purchasing data from the pharmaceutical company to see what impact my calculations could have.”
And what were the results?
Simon: “The results were breathtaking! With the purchasing plan of my algorithm, reductions of 8% in Total Cost of Ownership (basically all cost associated with the buying process) were suggested and I gained a lot of confidence in starting a business with that product.”
Was that the start for your own company?
Simon: “Not yet. While at the product side I was very confident, I first wanted to gain more business experience. I worked at a digital consulting company in Geneva for 3 years and met Arnaud, the brother of my wife, who had his own company at the time. After many discussions, brainstorming sessions and a big glass of wine, we both decided to step back from everything else and invest all our energy in GenLots. That was in May 2017, in the same week that my daughter was born.”
How did you finance yourself in the beginning?
Arnaud: “Initially, we invested some money ourselves and received another CHF 80.000 from business angels and H-Farm, an accelerator that hosted us for 4 months near Venice. The time in Italy was very fruitful, we lived in a beautiful environment with other startups in the countryside, developed the MVP of our software and received a lot of valuable feedback from industry experts. After we came back from Italy, we started working with the pharmaceutical company Merck to calculate better order plans for their factory in Aubonne, Switzerland. It felt good to finally use the algorithm with a real production facility.”
How does your algorithm, “the artificial supply chain brain” work?
Simon: “We use machine learning at the core of our technology. Our software processes a wide universe of possible solutions and isolates the absolute optimum in a few seconds. The underlying technology is called reinforcement learning, where the algorithm uses no historical data but learns from scratch by playing a game against itself. We set the game in a way that the algorithm gets rewarded when it chooses outcomes that reduce cost (e.g. for lower inventory, fewer order transactions or use of quantity discounts) and gets punished when production criteria are not met or cost increase. Our users eventually get a detailed plan that shows for each week how much of each material they have to order. The machine learning approach is foremost a massive shortcut for delivering a statistical optimum: calculation with traditional methods would take a standard server about 20.000 years.”
Note: In this blog post you can find out more in detail about the GenLots order planner software.
What kind of companies are using your software?
Arnaud: “Our product, the Order Planner and its new feature, the Opportunity-Dashboard is tailored for manufacturing companies with a purchasing volume of EUR 50+ Millions per year. Present purchasing approaches in these companies over-simplify the order process. Many order planners work with Excel-sheets, outdated scripts, and their own intuition when making important buying decisions. We help them by simply putting an intelligent layer on their ERP (enterprise resource planning) system.”
What do you want to achieve with your product?
Arnaud: “The main goal of GenLots is to offer a solution that makes industrial companies more productive. This comes without the need for reducing the workforce or relocating the factory but simply by using the power of calculation. But more generally speaking, we want to become a global player that transforms the whole supply chain scene itself. We think that the supply chain deserves more attention and tools designed specifically for their operators. The way companies are buying and shipping raw materials has not changed much over the past 30 years.”
We want to become a globally relevant player that transforms the whole supply chain process itself.
co-founder of GenLots
Why did you choose Lausanne as a location to grow your startup?
Simon: “First of all, we feel very much at home around Lake Geneva. Arnaud grew up here and I moved to Lausanne in 2006 for my studies and by now have my family and lots of friends here. Also, it’s great to have our office at the EPFL innovation park. EPFL (École Polytechnique Fédéral de Lausanne) is one of the leading research institutions worldwide for machine learning. This offers us a unique exchange with professors, excellent recruiting options and lots of exchange with other startups in the area.”
Arnaud: “We also have a broad customer reach from here. Switzerland is home to some of the leading industrial companies in the world with lots of food companies in Vaud and a strong pharmaceutical and chemistry sector around Basel. Including the South-German area, there’s also a huge amount of world-leading manufacturing firms. Our country is very compact and also home to many branches and buying centers of fortune 500 companies. Overall, we feel the location is perfect for our product. ”
Where do you see GenLots in 2 years?
Simon: “In 2 years, we want the GenLots software to be a household name in the supply chain scene with proven savings in our 4 core verticals. But most importantly, we want to have an outstanding team with fun people to work with across several offices in Europe.”