The solution is based on computer vision. This enhances the efficiency of the use of shelf space in retail stores. Technology improves the accuracy of assortment recognition to 99% and reduces the labor costs of sales personnel by 80%. Statistically, in order to increase turnover by 1% the on-shelf availability metric needs to be increased by 3%. Customers of Intelligence Retail have already achieved up to 15% on-shelf availability improvement.
Control of shelf product assortment, price and other information takes place online. Ailet was able to reduce the time taken to obtain an electronic report on a one shelf section audited up to ten seconds. The service recognizes up to 2.5 million images per month, and the number of users of the system, which exceeds 3,000. The image library contains about 100,000 units of food and other FMCG-items.
Ailet's product is used by Danone, Heineken, Perfetti Van Melle, Action group and others in CEE and EU, the Middle East and Latin America. Of the top 50 companies producing everyday goods, 12 are using the startup's solution, eight are piloting the service. In the case of Danone, the implementation period took one month. The same amount of time was required for tests. As a result, assortment recognition accuracy reached 95%.
The share of international project revenue is increasing also. By the end of 2021, Ailet plans to bring it up to 50%. An active expansion outside Russia will also require an increase in staff. At this time the company employs more than 40 people.
ExpoCapital Fund, as well as Aii Сorporation Oy and a few business angels financed the startup before this round. But Finsight Ventures, which invested in Badoo took the role of a Lead Investor. More recently, Intelligence Retail has signed a major roll-out contract for 12 countries in Latin America with leaders of the pharmaceutical industry. All the funds raised are intended for development, including in-store display recommendations services, based on big data.
Sergey Baramzin, CEO Ailet:
"We will direct the received funds to the product development, which implies image recognition enhancement, as well as full automation of the neural network training process. This will reduce the cost of running the solution and allow customers to start a mass use. In the future, the solution will not only collect information and measure the level of inconsistency between planograms and actual goods position's on the shelves but will also provide recommendations for an ideal out-of-shelf problem, building forecasts based on big data analysis."
Alexey Garyunov, Co-Founder of Finsight Ventures:
"We have been following the Ailet team for a year. This business shows a strong dynamic. There are a few similar international projects on the market. Products of competitive companies in the USA and Western Europe were created considering local labor market and retail context, which usually means they cannot compete in emerging markets where labor costs are scientifically lower. Basically, they focus on expanding the functionality using manual "repeated recognition". And it leads to significant delays. The Ailetl team offered an alternative model. It's a 100% automated real-time recognition.
Computer vision recognition technology is certainly the next step in robotics evolution. It is accompanied by increased efficiency and lower costs. We want to participate in this trend and hope that Ailet will be able to apply its expertise in various cases with insufficient digitalization, such as logistics fulfilment and production."
We will direct the funds received to the development of a product that involves improving the recognition accuracy, as well as the complete automation of the process of training neural networks for a new assortment. This will reduce the cost of launching the solution and open up the possibility of use not only to large manufacturers, but also begin mass use. In the future, the solution will not only collect information and measure the level of discrepancy between the planograms and the actual position on the shelf, but also give recommendations on how to build an ideal display of goods based on big data analysis."