A "robot" in custom orders/month two hundred thousand, 90% decline in demand and inventory 80% completely without human intervention

The retail future seems to have happened in a small office in Hamburg, Germany.

Otto - electronic commerce company, this is the biggest potential applications in artificial intelligence activities.

These "4 employees" is beyond your ability of people decided to micro and macro, enterprises in the colleague in the speed, accuracy, and efficiency of the highest in the company's history.

Big data (big data) and machine learning (machine learning) were applied for a period of time in the retail sector, the outstanding application is successful, Amason from the tech giant.

4 currently proven superior is huge in the section of the collection and analysis of data, from where you can put forward a hobby to predict customer products and blessed them P to improve customer satisfaction with the highest level of service.

The "Q" of the electronic commerce

However, the application of 4 was taken to another level in Otto. Learning in electronic commerce, data, and application of the machine, all the day-to-day decisions of large enterprises, not only in the customer service department, like Amazon.

Artificial intelligence is now responsible to reduce the cancellation rates, one is "Q" enable Otto to spend millions of euros a year of decision.

A disturbing mood to the e-commerce market, the trend is a customer to cancel the order or pay for the goods, if you receive the products, more than two days since I bought.

If these risks are delivered to the customer (including customer for products can be detected by a cheaper price, or decide to buy in the store, the customer is insolvent also that K, decided not to use more of the product, whether it is "God" does not spend a penny, when the cancellation and delivery, the opposite of Otto bear the freight and payment of goods are likely to be a waste of inventory V Because the product stage, through time.

In addition, customers receive the goods have hated many times, in a recent poll, most people buy things on the Internet people want in their own just bought a commodity only once.

Otto and because there are a lot of different suppliers, standing in a "dilemma" when customers place apart to wait until Otto: whether there is sufficient transport to reduce the time that the transportation cost Just like nor Otto packaging should be "hit" send something, you can avoid is cancelled by the client application.

In order to solve difficult mathematical problems, the Osman empire in the past, there is a plan to calculate the day's customers order products can be purchased in the future.

A department saw slow and low efficiency, a startup Otto investment technology blue yonder biggest potential applications in various enterprises of artificial intelligence.

Therefore, an intelligent algorithm for experimental analysis of the application of expert in Geneva's nuclear complex has entered the commercial activities every day, Otto.

Artificial intelligence analysis, more than 300 million orders in the past more than 200 different variables (e.g., keyword search history, shopping, and even the weather and Politics) measurement to estimate The catalog and quantity of goods project customers will buy in the future, "hand" before buying a step.

And effective "magic" of "4 employees"

Don't be ungrateful to bring Otto, artificial intelligence system, this can be said to be "magic": to accurately predict the variety and quantity to raise 90%