OSA Hybrid Platform

OSA Hybrid Platform is the provider of AI-based proactive scalable service for the FMCG sector, ensuring on-shelf availability of goods in retail stores in the real-time and in optimum quantity.
Our applicable AI engine processes data from nearly 100 sources. High accuracy of real-time out-of-shelf detection is achieved through the synergy of:
- Big data collection
- Real-time processing
- Advanced machine learning models
- Image recognition platform
- Deep expertise in FMCG-retail business and processes.
OSA HP team is based in 4 offices in Tallinn, Moscow, Kyiv and New York.

We have developed and successfully tested Optimal Shelf Availability (OSA) Hybrid Platform, a service capable of solving the out-of-stock issue on store shelves, which is recognized as the number one problem for today’s global retail, costing the industry 400 billion dollars each year in lost sales and another 100 billion dollars in product waste.

OSA Hybrid Platform is a profitable business established by a group of experts with backgrounds in business, data science and IT. Incorporated in Estonia, we operate in Russia, Ukraine and the US.

As of September 2017, we have contracts with 10 out of 20 largest consumer goods manufacturers operating in Russia. Among them: Coca-Cola, Danone, Efes, JTI, L'Oreal, Mars, PepsiCo, Unilever, SunInBev.

We have also signed up 8 out of 10 largest retailers in Russia: Auchan, Dixy, Magnit, Metro, X5 Retail Group and others, as well as one retail chain in Ukraine.

In the pipeline are the pilot tests in the US, France, China, Poland, and the Czech Republic.

It happened to all of us while shopping: instead of a particular product we were looking for, we found an empty shelf. The absence of consumer products on shelves in the retail stores due to the sell-out is the biggest problem the global retail faces today costing the industry $400 billion annually in lost sales. Out-of-stock ranges from 8% in the US to 21% in Russia.

The problem affects all parties.

Consumers: 55% of shoppers report negative shopping experience due to the failure to find at least one product they wanted to buy.

Consumer goods manufacturers: lose sales to competition when the retailers fail to display their products on shelf. They also lose the hard-gained customer loyalty, when shoppers are forced to switch to the competitor’s product due to their regular product being unavailable on shelf.

Retailers: lose shoppers to the competing retail chains and online retail. As a result, retailers report declining customer traffic and increasing costs of doing business. This pushes retailers to cut expenses, often downsizing the store staff, which only exacerbates the problem.

The situation is somewhat different for premium retailers: out-of-stocks are rare. Full shelves are maintained by charging higher prices, which allows retailers to accept higher product waste levels as they are fully covered by the extra margin they generate.

OSA Hybrid Platform is a service which solves the out-of-stock problem in the retail stores by generating and sending alerts in real-time to the store staff who are then directed by the system to fix the problem exactly where it occurs. Designed as a complete end-to-end solution, the Platform monitors the process from start to finish, finding and eliminating human error and process inefficiencies.

In order to generate real-time alerts in the store, our Applied AI Engineprocesses large amounts of data from more than 100 various sources. High accuracy of out-of-shelf detection is achieved thanks to:
1. Collection and processing of big data in real-time.
2. Advanced machine learning models.
3. Best-in-class image recognition platform.
4. Deep understanding of FMCG and retail business and all the processes involved.

Blue Yonder
Next Orbit
Simbe Robotics

The Platform's unique advantages are:
- AI-driven high-accuracy real-time alerts as soon as the particular product runs out of shelve
- The entire process of solving the out-of-shelf issue is controlled by AI
- A business model stimulating retail turnover growth while optimizing the manufacturer's costs and being free for the retailer

The pilot demonstrated the following results:
1. 70% accuracy of real-time alerts.
2. Sales growth of 5.4% for tested categories.
3. ROI in the range of 150% to 450% depending on the product category.
After the success of the pilot test, we have been strengthening the Platform with image recognition technology, which has increased our alert accuracy to 95% and offered wider functionality.

OSA Hybrid Platform has a unique business model based on win-win approach.

The supplier pays for the license, provides personnel to service the alerts, and helps us train our machine learning algorithms for a particular product category.

The retailer shares its data, performs regular inventory to optimize stock, and customizes its IT system to interact with our Platform.

1. IT and data science development
2. Master data catalogue
3. Image recognition model catalogue
4. Blockchain technology development and maintenance
5. Machine learning algorythms for the supply chain
6. Scaling business internationally

To be discussed during one-to-one meetings.

https://drive.google.com/open?id=0B5FxGKCftbvfaU4xenE0MDdpSzg