CUJO AI targets 'digital life protection' for service providers and their customers

With more connected IoT devices, such as locks, lights, thermostats and cameras on home networks, there's an increased need for better security.

CUJO AI is tackling the "digital life protection" problem through its artificial intelligence-based (AI) solutions that are in use by network service providers and their customers. To date, CUJO AI has announced two major partnerships with Comcast and Charter Communications, but CUJO AI President Andrea Peiro said his company has additional deployments in place in the U.S., Asia and Europe.

"So, CUJO AI is fundamentally an AI company that uses artificial intelligence and machine learning, specifically with deep learning algorithms, to address the challenge of providing network operators, and only network operators, with unified digital life protection solutions," Peiro said.

Peiro defined digital life protection solutions as a combination of network security, content control and digital control by parents as well as privacy and tracking protection. Peiro said CUJO AI has a two-pronged approach to its AI-based strategy.

"The first is empowering operators to provide their customers with value-added services that they can monetize directly in the domain of digital life protection, and the second part of our business is to provide insight and analytics and protection to the network operator itself," he said.

Contrary to traditional deep packet inspection (DPI) methods, CUJO AI is able to monitor the traffic flow of every device in a home by installing its private stack on gateways using open source frameworks. Peiro said DPI is becoming obsolete because of the encryption that it uses, and DPI slows down the speed of a connection.

"We combine our ability to look at data and how it flows in and out of the home and small business gateways that operators place in their customers' premises," Peiro said. "We can identify what devices generate this traffic, or what devices received the traffic and create AI-based maps, AI based networks, that map behavior exclusively looking at technical aspects of the network traffic, not the content itself," Peiro explained

CUJO AI's service provider customers are set up with private clouds that they operate themselves using CUJO AI's technology, Peiro said.

"Every single operator has its own level of privacy and security that they can control," Peiro said. "This information is aggregated in the cloud, and is used to build these models, as well as take action against threats, take action against potential content that needs to be prevented access to, or address privacy concerns."

For example, CUJO AI is able to detect when a cybersecurity attack has been spawned on a customer's video security camera. It also detects malware and attempted distributed denial of service attacks, and can see when bitcoin mining is being executed on a home network, according to Peiro.  

By blending data with the machine learning models and AI, service provider customers benefit from having a higher level of security in their home networks while companies like Charter and Comcast can use the information that's gathered for capacity planning.

RELATED: Organizations are starting to embrace AI

Peiro said that other startup companies are taking the same approach as CUJO AI, but they currently don't have the scale that his company has. Traditional security specialists, such as Symantec and McAfee, are also seeking to play a bigger role in home security

"There are some of the incumbents, the likes of Symantec or McAfee, that are looking at taking their traditional approach and scaling it to the operator level," Peiro said. "But what we're finding is that the ones that are more successful are the ones that are following us in the application of machine learning for mapping data patterns and movement behaviors.

"So, there is a whole roster of emerging small companies that have had more or less success. We are by far the largest from an installed base point of view, and the number of devices under our management."

To date, CUJO AI has close to half a billion devices under its management across 20 million protected gateways.


Mountain View, Calif.-based CUJO AI has about 150 employees in the U.S., Asia and Europe to support its global deployments. The privately-held company has had several rounds of funding, but Peiro didn't provide any additional details.

Peiro joined CUJO AI in February after previously serving as general manager and vice president of Comcast's advanced products group. While he was at Comcast, Peiro spent four years on the development of Comcast's xFI home networking management platform. He joined Comcast when Comcast bought PowerCloud Systems in 2014. Peiro co-founded PowerCloud Systems while also serving as CTO and product officer.

CUJO AI has been active in open source communities that focus on devices in a home. CUJO's software is integrated with RDK-B, which is the broadband version of the Reference Design Kit that is overseen by Comcast, Charter Communications and Liberty Global.  RDK-B is an open source initiative that standardizes functionalities for broadband devices, such as gateways, to speed up deployments to large customer bases.

Peiro said CUJO AI is broadening its portfolio on the privacy protection side of its business, and CUJO also sees an opportunity in 5G once it scales past the current "one-off" deployments by telcos. Peiro said there will be a convergence of 5G and customer premise equipment.

"So the convergence of in home, outside of the home, in business outside of the business, type of services to become independent from the transport is inevitable," he said. "We do have a value proposition in the 5G space that allows us to deliver services in the customer premise through cable or DSL gateways, and unify those value propositions when the customer shifts outside the home on a 5G network. So we are now working with some individual customers to deploy some of this."