Hi, tell us about your current role and the team/technology you handle at FortressIQ.
Hi, I’m the founder and CEO of FortressIQ. We’ve built a team of engineers and data scientists that build AI to help organizations to understand their processes. We believe that you can’t improve tomorrow, unless you understand today. Our teams work across multiple areas of AI, with a focus on Computer Vision, Natural Language Processing and Sequence Modelling.
What is FortressIQ and how it fits into the modern enterprise Analytics landscape?
One year after going public in 2005, Google acquired a small software company called Urchin. That acquisition became the foundation of Google Analytics. Google figured out early on that to advance user experience on the web, you’d need the ability to understand how users interact with web properties. Those abilities never made it to the enterprise, and this is one of the reasons why UX in the enterprise has lagged behind consumer UX. At FortressIQ, we’re bridging that gap, by providing an easy to use platform that can track the “journey” of different items like purchase orders, service tickets, and expense reports. By providing this new category of end to end process data, we empower organizations to understanding and improve their processes. We’re building the System of Record for Work within an organization, that can be leveraged across all of their analytics investments.
How do you see the growing investments into AI ML and analytics platforms transforming product-based business teams, especially in the Cloud and Security domains?
We’re big believers in the concept of augmented intelligence, and we see that as the bedrock of how these teams are leveraging AI. For example, dealing with the signal to noise ratio in alerts is the bane of a SecOps professional’s existence. AI is adding a ton of value to that problem, reducing that noise, and allowing humans to focus their efforts in a much more targeted manner.
The answer is quickly becoming not if AI should be integrated, but where. It comes back to a couple of central themes. Data growth is exponential. We don’t have the notion of a perimeter to monitor now, we need to monitor everything. There’s a great deal of work going on in unsupervised learning, which forms the basis of anomaly detection, which most of the ML based alerting systems rely on.
A great example of this is a new algorithm that just came out of academia, Microcluster-Based Detector of Anomalies in Edge Streams or MIDAS. This algorithm performs anomaly detection on graphs, which opens up a whole new area of exploration, and is more aligned to how sophisticated attacks are occurring. It’s also wicked fast, and more accurate than traditional outlier detection algorithms. Amazon also jumped in to new alerting strategies with an implementation of the Random Cut Forest algorithm in their Open Distro-based version of Elastisearch. New tools like these leverage cutting edge ML to help security teams deal with new types of threats.
Tell us about the current trends and technologies in:•
AI-based analytics – I really like what I am seeing in making analytics more accessible, by leveraging AI in natural language interfaces to analytics. I think we’ll see this in everything from corporate chatbots to Alexa. We’re also seeing a lot of activity around the idea that “Jupyter is the new Excel”. The plugin ecosystem to make Jupyter easier to use and add different types of functionality is rapidly expanding, and I see notebooks expanding past the data science teams and into the world of business analysts.
RPA – The biggest trend here is that enterprises are demanding more. Simple RPA delivers value, but it’s not moving the needle. With the continued commoditization of the bot itself, everyone is looking at how they can package higher level services with the bot to deliver increased value.
Cognitive Learning – Solving the data problem, is where the most exciting work is. Traditional ML models require mountains of data to produce results. Getting that data is becoming harder and harder. There are a myriad of ways that practitioners are looking at solving the problem with great work going on in areas like Self Supervised Learning, Self Supervised Contrastive Learning, and Synthetic data. Any time you can have AI train itself, you’ll reap benefits.
Hear it from the Pro: Will AI-as-a-service become the cornerstone of all Enterprise IT stacks?
Simply put, absolutely. Most companies shouldn’t be building their own hardware, actually most companies shouldn’t even be building their own software. Building differentiated, scalable, resilient, and ethical AI is very, very hard. I don’t think most companies can justify the sustained investment in “ground up development” when their competition will be able to leverage off the shelf AI-as-a-service and innovate quickly.
Enterprise RPA is a booming industry. However, COVID-19 may have impacted the run. What is it like to continue innovations and product enhancements at FortressIQ during the pandemic?
They say that necessity is the mother of invention. In a scenario like today’s economic climate this it is more important than ever to get back to the basics of what makes a well-run startup so formidable. Listening to customers, and incorporating their feedback at a breakneck pace while delivering value in every interaction. One of the major themes we are seeing is how can you ensure that customer can get maximum value from your offering without the traditional enterprise style roll out and training. In this distributed environment we can’t expect the same level of time and effort to be invested to receive value from new products. That’s forcing us to think of simpler ways for users to interact with our software, realizing they just don’t have the time or support infrastructure to deal with obtuse and clunky software.
Thank you, Pankaj! That was fun and we hope to see you back on AiThority.com soon.
Pankaj passionately believes in AI’s ability to improve the human experience, and looks forward to a world bereft of mundane tasks, where people get to focus on being their best selves.
Before founding FortressIQ, he was the CEO of ThirdPillar systems, where he designed and delivered lending platforms for commercial banks, processing billions of dollars in transactions. ThirdPillar was acquired by Genpact, the world’s largest BPO company, where Pankaj went on to hold serveral senior management positions, culminating as a founding member of the company’s AI lab in Palo Alto, California.
FortressIQ is the creator of a cognitive automation platform that powers and accelerates digital transformation through imitation learning. Using an innovative type of AI that combines computer vision, natural language and sequence modeling, FortressIQ learns how a business functions through live activity analysis. By radically lowering the cost, effort and time required to document business processes, enterprise organizations can use FortressIQ’s platform to quickly gain the insights necessary to improve business operations and implement automation.
The FortressIQ platform automatically discovers, maps, and documents all the digital processes executed by an organization’s workforce. From web-based apps, to legacy systems, virtual desktops & green screens, we eliminate the need for APIs and access to event logs, requiring no configuration and zero integration, delivering a time to value in weeks. With AI-driven technology, the discovery method generates process documentation automatically, allowing customers to map processes in record time, and transparently providing the most holistic view of their current state digital operations.