PROBLEM STATEMENT
Sources of cyber attacks
-
Internet -
Corporate mail -
USB devices -
Mobile devices
According to IBM and the Ponemon Institute,
the average cost of a data breach
in the financial sector above
$ 5 000 000
The financial sector is the most
popular target
of cyber attacks with an average loss above
of cyber attacks with an average loss above
$ 18 000 000
Threats
-
Customer
base loss -
Bank website
spoofing -
Financial
asset theft -
Intellectual
Property leakage
RISK OVERVIEW
-
Payment and billing systems
-
National Treasuries
-
Crypto trading platforms
-
Reserve Banks
-
Commercial banks
Potential Damage
-
Extortion of funds
on a large scale -
Theft of customer
and employee data -
Theft of finances and
loss of reputation -
Loss of clients
and partners
Our Solution
FV-SANDBOX, a comprehensive tool for protection of
bank services, employees and business processes developed to protect banks from cyberattacks
How it works
- Protection against phishing and spam, which are the most popular attack vectors
- Control of files downloaded by employees from the Internet
- Monitoring of USB devices connected
to workplaces
- Security scanning of business emails
- Safeguarding of remote access
Purpose
- Mitigation of operational and credit risks
- Protection of personal data of employees and private information of clients
- Countering financial fraud
- Enhanced training for IT
security specialists
scheme of work
Features
Own database of signatures
Bot system for collecting IoC
Integration with deception platform
High speed up to 250,000 files per hour
Two systems in one antivirus
multiscanner and sandbox
multiscanner and sandbox
Machine learning models with
the possibility of retraining
the possibility of retraining
Spam checking
A large number of machine
learning models based on new
transformer algorithms
learning models based on new
transformer algorithms
The possibility of automatic
and manual additional training
of models on user data
is provided
and manual additional training
of models on user data
is provided
Multilingual with support
for more than 15 languages —
Arabic, Chinese, English,
Spanish, etc.
for more than 15 languages —
Arabic, Chinese, English,
Spanish, etc.
Protection of models
from poisoning based
on the control dataset
from poisoning based
on the control dataset
Machine learning
Using machine learning models for various
file types with the possibility of automatic retraining
file types with the possibility of automatic retraining
Spam
MS Office
Phishing
DLL
PDF
APK
EXE
ELF
Extreme Gradient Boost
VGG, NasNet, EfficientNet
Catboost
Light Gradient Boosting
RandomForest Classifier