Open Data
matters now

The customers:

Public organizations, government agencies.

The problems we solve:

Open sources contain a huge amount of data, which allows us to solve a wide range of socially significant problems. We distinguish two major areas of such tasks: an in-depth study of certain qualities of a person (for example, suicidal tendencies or pedophilia) and anomalies in the informational space (for example, identifying bots and fake news).

The benefit:

Providing additional information (in a real-time mode, if necessary) which significantly increases the efficiency of solving the tasks faced by public organizations and government agencies.

The format:

Access to the personal account with a variety of analytical tools. Information in the personal account is regularly updated.

The mechanism:

To study human qualities, we combine the results of international studies and the expert opinions of our psychologists, imposing it on the volumes of available open data. To study anomalies in the informational space, we have designed our own advanced neural networks and other algorithms for analyzing textual information in Russian and Ukrainian languages. For each task, we develop a set of analytical tools to obtain the maximum number of insights.

The customers:

Online and offline retails, banks, online business, etc.

The problems we solve:

Addressing a wide range of tasks related to understanding behavior and preferences of the target audience. This includes a detailed portrait of the target audience, identification of the key segments, studying the psychotype and the needs of the audience, perception of information by the audience (text and visual), communication channels with the audience, differences between the various audiences of the company, and others.

The benefit:

We show the new vision for the business based on the external open data to get away from the one-sided picture, which appears if only internal data is used. Combining external and internal data provides additional strategic insights for business development.

The format:

The results are provided in the form of a pdf report.

The mechanism:

We analyze digital traces of the target audience in the open sources, considering only publicly available information. Using machine learning algorithms, we clear and structure information. We also develop our own neural networks and other algorithms for cleaning and structuring text information and images. For each task, we develop a set of analytical tools to get the maximum number of insights.

The customers:

Online and offline retail, banks, telecom, online business, etc.

The problems we solve:

Building predictive models in cases where internal data is insufficient, including up-sell, cross-sell, next-sell models, forecasting fraud, personalized loyalty programs, advertising offers, and others.

The benefit:

For example, for most next-sell companies, a model based on internal data can be built for 5-10% of all customers (with adequate accuracy), as the company has too little interaction history for the rest. Adding external data to the internal one increases the percentage of coverage in many times.

The format:

SaaS solution: model training and data markup on the initial stage with the results being further provided via api in real time.

The mechanism:

Combining internal company data with customer data from open sources. Using the training sample, we build models that predict the necessary parameters. For example, having an initial sample of sales, we combine internal data with external data and build a model that finds the relationship between the goods sold and the type of customer. Then, with the help of the trained model, all other clients are marked.