Investigate, prevent and mitigate risks of riots,
terrorism, disasters and crimes through
a country-wide geo analytical AI

Use cases:

National security

Artemida detects abnormal crowd density at any location and alerts
the authorities in case of riots, violence, and disturbance of peace. It allows
to automatically manage the network services, downgrade or disable internet
connection, and extract the exact number of MSISDNs participating in riot

Real-time data

Moreover it provides accurate data (location of all mobile phones is updated every 15 minutes)

Tracking roamers

Subscribers' movements and historical data about their location

Protection of critical infrastrucuture

Geofencing allows to set up rules and the management, to get data in time

Crime investigation

Remove the noise.

Artemida algorithms classify MSISDNs that belong/live in the specific district or an area in the city

Unknown phone detection.

Identify the unknown number travelling with your suspect

Identifying the suspects by their route.

Artemida finds all MSISDNs on the target’s given route

SIM card swap detection.

Receive information on all previous and current SIM cards installed on the phone

Continue tracking the same phone even if the SIM card is swapped

Riots, violence, and radicalism.

Find all the participants in given locations and time

Public safety

Getting the exact number of people affected by
In cases of district, city, or countrywide
emergency, Artemida allows us to know how
many people were present when disaster

Emergency alerts. Send custiomized messages to subscribers in the area of an emergency. It helps emergency services to deliver the crucial instructions to the people affected by the disaster in time.

Mark and control the spread Artemida detects the location and density of mobile phones of infected population.

How it works

Artemida uses information from LBS module, which pages all the mobile devices every 15 minutes and stores the data (for 2 years)

We have developed complex ML algorithms that solved following problems:

  • Implementing AI significantly increased the accuracy
    of identifying the number of MSISDNs in any specific area
  • Allowed to decrease the computing time without
    losing the accuracy of MSISDN location
  • Our ML reduced the cost of the solution
    by 200 times in comparison to other rivals
  • We have developed smart visualization
    dividing the whole country, city by hexagons
    which helped to increase the accuracy
    of the density of the crowd


  • Retrospective analysis —
    determining when and where
    crowd began to congregate,
    how it moved and scaled
  • Look alike algorithms —
    approximate identification
    of similar persons by matching behavior
  • Profiling - identifying home
    and work places and more
    based on the behavior pattern