Information security service for risk analysis

Task: develop a web service for risk analysis, data visualization, and tabular data analytics using pre-built mathematical models

Duration – 1 month
Team – 3 members
Technologies – FastAPI, Postgres, Streamlit

The client is one of the largest Russian retailers, requiring software to calculate risks and track incident counts for their commercial operations. Previously, the client relied on Excel combined with the ModelRisk plugin for analytics. However, this solution was slow and costly, prompting the search for a simpler web-based system.

We proposed developing a product using Streamlit, a Python library that enables the creation of web applications for machine learning, data visualization, and other tasks. Streamlit provides a simple and intuitive way to build user interfaces.
This library also allows for easy and informative data presentation, making it possible to quickly create interactive dashboards for data visualization, which makes the data analysis process more accessible and understandable.

After selecting the technology stack, we moved on to implementation. Our task was to create a web service for forecasting the number of incidents and simulating data from ModelRisk. During development, we relied on pre-existing mathematical models provided by the client, as well as the structure of the Excel file the client previously used for calculations.
The parameters defined in the model are calculated based on probability distribution functions. Then, using Monte Carlo simulation, possible scenarios are generated for the model’s uncertain parameters.
The output of the model is an estimation of the number of incidents on a yearly and monthly basis for future periods, taking into account the selection of distributions based on uploaded cybersecurity incident statistics.

Ultimately, the product addresses even more business needs than initially anticipated. After preparing documentation that includes the main commands, a detailed description of the solution, and various ways to interact with the system, we delivered the project to the client.

Currently, the service operates autonomously within the client’s internal network. We continue collaborating with the client, further developing their internal cybersecurity services ecosystem.