Guaranteeing Privacy in a Time of Unprecedented Uncertainty

Three years ago, we founded FortressIQ with the idea that by leveraging computer vision, we could radically transform the way organizations understand their processes, thus accelerating their transformation. A little over a year later, we released our process intelligence platform, which has been deployed across Global 2000 companies in diverse industry segments. At the same time, the pace at which computer vision has been advancing is nothing short of breathtaking, but with great power (AI) comes greater responsibility. At FortressIQ, we believe that if we invest tens of thousands of hours in improving computer vision, we also have an ethical and moral responsibility to address use cases that can abuse those same advances. With that in mind, today we’re releasing Privacy Enhanced Gateway (PEG), the world’s first adaptive and learning computer vision-based firewall. And to stand unambiguous in the belief that privacy is a human right and that organizations who create technology that has the potential for abuse have an obligation to provide solutions to that abuse, we are making Privacy Enhanced Gateway a standard part of our platform at no additional cost.

With the adoption of FortressIQ across major global enterprises collectively employing over 5 million people, there’s proof in the industry that computer vision-based process intelligence can deliver unprecedented levels of information, creating an entirely new category of data and insights. The need for process intelligence has only become more critical with current workforce upheaval caused by work from home orders, global supply chain disruptions, and ongoing uncertainty. Part of our mission is to expand the usage of process intelligence to improve outcomes across the enterprise and to mitigate external forces at play. But, to attain the level of ubiquity that we’re seeking, we can’t ignore the security and privacy concerns associated with large-scale deployment of computer vision-based applications.

The scale of a challenge created by AI can only be solved with AI. This is why we took our expertise in computer vision (object detection, scene text detection, and recognition), natural language understanding, and data mining and created PEG, the world’s first learning and adaptive computer vision-based firewall.

A core component of our Process Intelligence Platform is the software-based sensors that are deployed on an end user’s desktop or laptop. These sensors tap into the video card of the machine and create a low bandwidth video of all screen activity, which streams to our cloud, where we convert the unstructured video into a structured log file. While we are HIPPA, SOC II Type 2, and ISO 27001 compliant, the idea of streaming a user’s activity (with the associated data) to our cloud, proves to be challenging.

Privacy Enhanced Gateway essentially takes some of the core AI from our platform and allows it to run at the edge under customer control. It is packaged as a virtual appliance that can be deployed on-premise or within a customer’s VPC on the cloud provider of their choice. Any sensors deployed then connect to their private gateway, which redacts all customer information from the video stream before sending the data onto us.

The core of this technology is a joint AI model which has learned the individual types of objects which compose a software screen, and the meanings of their labels and values. We can determine every control, like a drop-down list box or radio button, the label for that control, along with the data within that control. We then can redact the data leaving only the empty software screen remaining, thus alleviating the challenge of protecting sensitive data.

Using PEG, our customers get the best of both worlds. They can leverage the scale, cost savings, and rapid advancement of our cloud-based platform with the security and privacy of an on-premise deployment.

PEG takes us one step closer to unlocking the limitless potential of the global workforce by accelerating the responsible and ethical use of AI in the enterprise. We look forward to partnering with our customers to provide an unprecedented level of insights while at the same time proactively protecting the security of their data and the privacy of their employees. We’ve shown how computer vision-based AI can co-exist with privacy, and we invite other process intelligence vendors to join us in delivering responsible AI-based solutions.