Darktrace

The typical use case for Darktrace  is for threat vector scanning, detecting any unusual activity, and anomaly detection. Apart from that, it is very helpful in incident response.

The features I find most effective in Darktrace  include anomaly detection. The machine learning model provides accurate alerts after the learning period of 1 or 2 weeks, especially for network anomalies or something that the user is trying to access, which can include trying to visit unknown sites or botnets, and those things get detected and represented in a very good dashboard.

Darktrace positively impacts my organization by enhancing threat hunting, particularly in east-west traffic within the same subnet. Previously, we only used traditional firewalls that cannot catch this lateral traffic. After deploying Darktrace, we gain insights into machine-to-machine communication, which adds more value to the organization and is especially beneficial for the SOC team.

In terms of improvement for Darktrace, pricing is the main concern. Pricing bothers me and this is one of the major factors when choosing a solution. When we get feedback from customers, that's the only felt need. When we factor in Darktrace, we do it only limited. We put it on where the perimeters and connections are, but still, some gray areas are left out, especially if we have multiple branches. We need Darktrace on each branch to get the data out, and I suggest having some kind of a centralized product that gets data from multiple sources to aggregate and provide the data.

I have been familiar with Darktrace for the last 5 to 6 years.

In terms of the speed and effectiveness of Darktrace's automatic response, it gives clear alerts whenever anomalies happen on the network, enabling us to catch them on the fly. However, some of the rules generate false positives, especially with system calls, which get incorrectly marked as anomalies. These are actually system call integrations that need fine-tuning based on our environment integrations.

Regarding Darktrace's capability to adapt and recognize abnormal activities through machine learning and AI, sometimes a password expiration prompts the user to connect to different sources to get the new password changed. During that time, it picks this up as abnormal activity when connecting to LDAP during off-business hours. This is an example of how it detects what it considers an anomaly, since user authentication typically happens during business hours.

Regarding overall stability, Darktrace is a stable product, and I have no complaints from customers wherever it is deployed.

While considering if Darktrace is scalable, I note that there are storage limitations, where the planned capacity can sometimes be overutilized. There is still a gap in terms of storage, and we are trying to figure out how to increase that capacity for regulated environments, which require data retention for 5 to 6 years.

I can rate Darktrace's technical support as one of the best products in the world. We have seen satisfaction reflected on our customers' faces after deployment when they start seeing the data and the dashboard, and they often express surprise at the network traffic visibility that Darktrace provides.

I would rate the technical support of Darktrace between 6 to 8, as the support is good and we receive timely assistance whenever we raise an issue.

Before working with Darktrace, I did not use any similar solution in the same category. Earlier, I was using something called decepters, and my organization may have explored different products, but I learned about network detection and response through Darktrace about 5 to 6 years ago.

Deploying Darktrace is quite easy and plug and play, wherein all we need is to put it in a data center, rack up, and do some switch configuration. The learning would take a week time, and once the data gets populated, we get a very good dashboard.

For deploying Darktrace, I would require 3 to 4 people. We would require a data center person to assist in racking and mounting this, and some network engineers would make this configuration to spend the data ports.

When considering return on investment for organizations using Darktrace, the disadvantage lies in having to use a physical appliance. Running a quick POC is not possible since the hardware has to be shipped from the UK or elsewhere, but other NDR solutions provide virtual appliances that can be deployed on virtualization servers to get up and running quickly.

In terms of setup and licensing costs, Darktrace is on the pricier side compared to similar solutions in the NDR market. Other NDR solutions are also on the higher side, but Darktrace stands out as a bit higher. Competitive pricing would certainly help me as a system integrator to convince customers.

I did not evaluate other options when looking into Darktrace, but some customer preferences led us to consider other NDR solutions, such as 40 NDR. Our customers had a Fortinet setup with various products, and they preferred the 40 NDR for proprietary visibility when collecting logs from Fortinet devices.

We are using the latest version of Darktrace. I have not used Darktrace's Enterprise Immune System. Antigone is the feature of Darktrace that we have recently experienced. At the moment, I have not encountered a situation where Darktrace's self-learning capabilities reduced the risk of data breaches, but it performs very effectively overall. It requires some time to adapt; initially, when we deploy, it takes weeks. On a scale of 1-10, I rate Darktrace a 9.

On-premises

Other

I am using Darktrace  for email security purposes.

The most beneficial feature in Darktrace  is identifying phishing emails with the help of the AI engine and machine learning. In case it does not identify something, we can automatically make Darktrace learn from selections and other functionalities.

Regarding the ROI, we have experienced a significant reduction in phishing emails and have utilized our time efficiently, resulting in approximately 70% ROI.

I feel that Darktrace could be improved, particularly in the support aspect which is currently very poor. We need to chase Darktrace instead of them being proactive with us.

The support is the main problem, though there are some other issues as.

I have been dealing with Darktrace for eight to nine months.

We have faced some integration issues when integrating with CrowdStrike, and we are still facing these issues because the support is very poor. Similarly, we need to integrate with our SIEM  solution, which is experiencing issues. We cannot proceed significantly because the support is not very active and sometimes gets delayed, so we need their engineering support and other assistance.

Regarding scalability, it is very stable, and it is 100% scalable similar to CrowdStrike.

If I were to rate support from 1 to 10, where one is bad and 10 is good, I would give them four points.

Neutral

The installation is quite plug and play.

Everything is totally handled by Darktrace, and I do not have to do anything for maintenance.

The pricing is affordable and not very expensive. If we rate it on a scale where one is cheap and 10 is expensive, it would be three points.

The Autonomous Response capability in Darktrace handles real cyber threats quite efficiently. The Autonomous Response is excellent at identifying phishing emails and suspicious emails accurately, and it automatically sends a response to users that certain emails were blocked by Darktrace, helping users identify whether it was done correctly or incorrectly. If it was done incorrectly, the user can submit a request, and we can perform human analysis and then add it to a whitelist or blacklist.

In terms of AI functionality, I have seen some AI integrations overall. Darktrace is completely designed based on AI and machine learning, making it very efficient in identifying suspicious behavior and suspicious emails.

We are using the Securonix SIEM  solution, and from ManageEngine, I use Help Desk and the Patch Manager .

On a scale from 1 to 10, I would rate Darktrace as six points.

Public Cloud

Microsoft Azure 

My current use case for Darktrace  is network detection and response.

The functions I find most valuable in Darktrace  are the AI analyst as well as the detection.The autonomous response capabilities of Darktrace are not crucial for me because it doesn't work in a network where there are no core switches. In a modern network, the autonomous response doesn't work, especially when sitting in a shared data center.If I'm running a traditional network where I am not in a shared data center with a layer two dedicated for my resources, then it can work for me. However, if I am in a data center where I don't have layer two, it becomes an issue because the autonomous response is reliant on sending spoofed TCP resets to my core switch to block traffic, which is a major issue.

I am uncertain what would make Darktrace better because of the autonomous response issue. In a shared environment, it doesn't work, and there are still some integration issues. They say they can integrate with most firewalls, but when we did an integration with Meraki MX  firewalls, that integration didn't work and still doesn't work to this day.

I've been working with Darktrace for more than five years.

For stability, I would rate Darktrace an eight out of ten.

For scalability, I would rate it an eight out of ten because they integrate with many technologies.

The technical support from Darktrace is very good, including support from their resellers. We worked with Grove, who are with 360 Integrity now, and they are quite good.

Positive

The initial setup process for Darktrace is complex because I need a network technician or engineer to configure the span port on my core switches, and I need assistance to choose which VLANs I want to ingest traffic from. Beyond that, everything is easy; the management of Darktrace is quite easy. Making exclusions is easy, and investigating within the platform is quite easy. However, the initial setup becomes complex due to the requirement of getting someone to create the span, and needing a dedicated span for Darktrace on the switch.

The main competitor of Darktrace at the moment, based on how long they've been in the market, would be Vectra. Vectra does a similar thing, but Darktrace would beat Vectra based on the algorithms that Darktrace is using because Darktrace's algorithms in terms of AI and ML are quite good.Cisco is also coming with some technologies such as Cisco Secure Analytics, network analytics, and Cisco Secure Cloud Analytics . That technology is quite good because if running mainly on Cisco, such as SD-WAN Meraki devices with remote workers using Cisco Umbrella  or AnyConnect, pulling data from them and pushing it into an XDR  or Secure Analytics can be quite effective, providing reachability, visibility, and scalability.

Regarding the number of IPs monitored using Darktrace, they provide licenses that allow monitoring of around 16,000 IP addresses, and they give a buffer. Monitoring is possible for as many devices as long as they are in line with the traffic. However, devices that are not in line with the spanning won't be visible, which means there is no visibility.My recommendation for Darktrace would be based on the kind of network someone is running. If someone is running a network where they have dedicated layer two switches, I would most definitely recommend Darktrace as it's a good product. However, if someone is running in a shared environment where they share the layer two with other customers in that data center, then Darktrace wouldn't be suitable.For an overall rating, I would give Darktrace an eight out of ten.

On-premises

Other

I use Darktrace  for incident response and detection within my organization.

The AI analysis and AI investigation features are incredibly effective. I do not need to manually process incidents as Darktrace  provides an incident summary, potential detection paths, and other details, all exportable with just a click. The tool is very powerful and saves a lot of time. The autonomous response technology eliminates the need for human intervention by automatically handling incidents even during off hours.

Updates keep coming, which is great, but I prefer a unified UI experience. The intelligence section and the incident view should be seamlessly connected in one view to avoid jumping between pages. This integration would make incident management more intuitive.

I have been working with Darktrace for one year and two months.

The technical support is excellent. Any queries I have logged are responded to efficiently, and my cases are promptly managed and resolved.

Positive

Before using Darktrace, I worked with NetWitness and RSA NetWitness.

The licensing cost is approximately eight dollars a year. The cost is reasonable considering the unique capabilities and features Darktrace provides. The support is included, and any issues I have are addressed through logged cases.

I would 100% recommend Darktrace. The product is autonomous, detecting and preventing threats effectively, unlike many competitors that are stuck only at detection. The visuals and the conceptualized views for connections greatly assist in threat analysis. My rating is eight out of ten.

On-premises

We use Darktrace  for network detection and response, specifically for monitoring the network for any malicious activities, anomalies, and lateral movements.

The autonomous mode, which is the Antigena AI response, is particularly valuable. It is capable of responding to lateral movement and ransomware deployment within environments where there is data exfiltration. For example, if more than 2.5 gigabytes of data have been pulled in a few minutes, it engages by blocking for one-hour intervals, alerts, and extends the block until it goes into full isolation if the violation continues.

I have observed a product called LinkShadow, which has more modules of integration and consolidates everything into one dashboard, such as authentication monitoring. Darktrace  could improve by integrating with email security gateways like Mimecast or Ironscales. In LinkShadow, this is referred to as mesh technology. Additionally, the Darktrace dashboards are not easy to navigate until one becomes familiar with them.

I have used this solution for the past four years.

There was no mention of specific deployment issues, but the deployment took within a week.

I would rate it a ten. The only time it has been unstable was due to network links going down where other probes were deployed. The appliance itself has never let me down.

I would rate the scalability a ten.

I would rate it a nine. The challenge lies in waiting for a response after logging a ticket. Sometimes it requires back-and-forth responses before having an actual conversation with technical consultants via calls or video conferencing.

Positive

Before using Darktrace, we never had a different solution. We did not have a network detection and response tool.

The initial setup is straightforward. The machine needs time to learn the environment before entering active mode. It needs to learn traffic flows and what is considered normal before activating full autonomous mode.

We worked directly with Darktrace for the implementation.

The pricing is costly in USD, and they charge based on device counts. Other network detection and response solutions use module-based licensing with unlimited device connections, but Darktrace charges based on the number of devices ingesting data.

We tried LinkShadow and considered Cisco Secure Network Analytics , but pricing was an issue there, so we never ran a proof of concept for it.

I would recommend Darktrace and rate it a nine. I have not evaluated many other competitors, other than LinkShadow, considering our budget constraints.

On-premises