Our company offers technological solutions for your business in six key areas

Content Management

Objective: a system that allows uploading, storing, processing and broadcasting any type of media content (images, video, audio)

  • Grabbing data from any source to store it in a generic format for further processing
  • Universal conveyor that allows for the necessary preprocessing of the media content
  • Moderation, both manual and automatic
  • Loading the content in CDN, content management (modification, deletion), content migration to other repositories

Video processing, storage, and broadcast

Objective: geo-distributed video content broadcast system

  • Downloading videos from various sources, both from simple repositories and sites that don’t provide for easy video downloading
  • Multithread, multi-server video conversion
  • Uploading the finished videos in any convenient repository
  • Player that supports video streaming via HLS or DASH

Image processing analysis and classification system

Objective: image processing and classification with the addition of useful metadata

  • Downloading images for further processing and queuing for distributed image analysis
  • Image analysis using computer vision libraries with a view to:
    • Calculate coordinates for cropping unwanted elements (watermarks, strips)
    • Calculate settings for graphics filters
    • Carry out assessment of image quality
  • Image analysis by neural networks for the purpose of classification and selection of meaningful signs
  • Uploading pictures to any required storage
  • Getting images from video by using an algorithm designed to calculate the most valuable and meaningful frames
  • Processing multiple frames for better image analysis

User behavior evaluation system

Objective: generating a user profile with the possibility of deeper user behavior analysis

  • User registration in the system, user labeling, parameter identification and segmentation, fingerprint generation
  • Recording all user actions in a database
  • Building a model user behavior algorithm and adjusting a model microsegment behavior algorithm
  • Calculating the actual user value
  • Adding information on this user and microsegment to the statistics database system
  • Implementing deep user behavior analytics capabilities by analyzing segments and microsegment behavior data
  • Creating predictive resource development models using data mining and machine learning

Automatic traffic purchase and distribute system

Objective: implementing all phases of attracting, evaluating the quality, distributing and determining the actual effectiveness of users for the resource

  • Connecting to traffic purchase systems from different sources using the RTB protocol:
    • Obtaining data on user parameters according to the RTB protocol
    • User segmentation based on these settings
    • Forecasting the user value and calculating a fair price based on the segmentation used
  • Making a decision on purchasing and providing bid response
  • Adding a user to the system after a successful “purchase” via the RTB protocol, order-based purchase or direct purchase through internal system
  • User registration in the system, user labeling
  • Definition of user parameters:
    • GeoIP (country, state, City, Timezone, Locale)
    • Connection type
    • Isp
    • Referrer
    • Device (Type, Manufacturer, Model, OS, Browser, Screen size and other)
    • Proxy
    • Cookies, etc.
  • Assigning a specific microsegment to a user depending on the settings
  • Generating a user model fingerprint
  • Generating a user value forecast based on the statistics on similar user and microsegmentation types
  • Identifying the best landing page with the maximum predictive value for a given user
  • Redirecting a user to the selected landing page
  • Calculating the actual user value based on user behavior results
  • Recalculating the value of the microsegment and dependent segments
  • Recording all user behavior indicators in the statistics database