Works based on computer vision technology, natural language processing and machine learning. My name is Anna Polishchuk, I am 35 years old. In January 2016 we partner with Yury Mukhin has launched a mobile marketplace Lalafo.
Lalafo is a mobile application for sales b/u-things and search real estate, jobs and services. The service uses technology computer vision and NLP (natural language processing) based on neural networks. For training neural networks Lalafo uses proprietary algorithms and a dataset of 14 million photographs of products, published and described by users of the service.
Features: Since 2010, I have worked in the London office of the service Slando, who was responsible for the development of the product in 11 countries. Before entering Slando to Naspers media company, I held the position of commercial Director.
After the deal with Naspers, I became the Executive Director of Slando and was responsible for the development of the service in four countries, including in Russia. At the same time, Yuri Mukhin worked in company marketing Director. By 2013 Slando reached a monthly audience of 27 million users in four countries.
We know that in a month only 1.5% of the global Internet audience sells on the marketplace and services announcements, and about 10% — buy. However, in every home there are thousands of goods that people dont use. Billions of dollars that are in garages, cottages and balconies.
We have identified barriers that impede trade between private people. Now the volume of the global market classifides approaching $100 billion. If you remove barriers, it can increase ten times.
We have created Lalafo to remove these barriers with the help of artificial intelligence. Computer vision, natural language processing and machine learning. We apply our own algorithms for training neural networks. They extract from each photo set of characteristics clusterize photo in multidimensional spaces.
For some categories, such as jobs and services, the text is quite important. Our algorithms for natural language processing can now define the category of jobs with an accuracy of 96% in almost any language. Lalafo now working in Serbia, Greece, Azerbaijan and Kyrgyzstan.
For the first year, the service reached a monthly audience of 3 million users. At the same time he processed 15 million photographs of products. Uploaded, categorized, described by the users and approved by moderators. When a new object is placed on Lalafo, the algorithm instantly retrieves its characteristics.
They are used to determine the correct categories and calculate rates for a specific geographic region. Based on these data, a “smart” algorithm finds the most suitable buyers, the interests of which he knows. The app is available for smartphone users on iOS and Android. Want to get the word out and to tell about your startup.
Welcome to rostrum.