The main challenge that we wanted to solve with this solution was the email-based attacks. They are a major threat for our staff and customers. So, we are using this solution for protection from threats and remediation of phishing emails. We also use it for simulation to train our staff to recognize a phishing email. For those users who click on the phishing email, we provide the training content to improve their awareness.
Its artificial intelligence and machine learning capabilities are really important and helpful in saving time. It saves a lot of hours, and we don't have to do a manual review of each incident or email. Without IRONSCALES, we will have to rely on our staff to report emails that might be malicious, but you can't always depend on the staff to do that.
It is an important part of our security infrastructure. It gives us confidence that we are protected from email threats.
It allows us to report an email as spam. If our staff suspects an email to be malicious, they can report it as spam.
It is helpful in assessing the awareness level of our staff. We compare the metrics of every new phishing campaign with the metrics of previous campaigns. We also determine if the awareness training we are conducting for our staff is effective. We can see all the data. If a lot of people have clicked on the phishing email, we get to know that we need to step up our efforts on phishing awareness.
Its automated detection and response capabilities save a lot of time. We are a financial institution, and we get quite a lot of emails. If someone has to manually review and investigate each alert or email event, it would be a full-time job. On a daily basis, it would take more than three hours of analysis work. A financial institution like ours receives quite a lot of emails from customers and internal staff, and IRONSCALES saves a lot of our time. We don't have to dedicate someone to manually review alerts or emails.
The time that we have saved by not having to manually review each incident can now be used to focus on other tasks.
It has also helped reduce the number of people or analysts needed to deal with and respond to email incidents.
It allows us to customize the automated detection and response capabilities. Each email that IRONSCALES detects has a confidence level rating. For example, it has 90% or 70% confidence in a phishing email. We have the ability to configure the threshold for automatic remediation. For example, we can say that for emails with a 70% or higher confidence level, it can automatically remediate, but for emails with a 69% confidence level, it should raise an alert, and we'll manually investigate. It gives us the ability to raise or reduce the threshold.