As the world continues to fight COVID in this 10-month long battle, we all have learnt to survive, adapt and sometimes even improvise beyond all means. For several industries, it was a moment of truth – to digitize or perish. The transition to Industry 4.0 was a sparingly discussed concept until March, following which its application has skyrocketed. Almost every major industry faced disruptions in supply chain, vendor management and other critical processes earlier this year. It was more important than ever for companies to embark on their digitization journey, lest they get left behind.

For Dr Soudip Roy Chowdhury, CEO of Eugenie.ai, it was a moment of reckoning and reinforcement of his belief that 2020 would be an important year for edge computing and Industry 4.0. An unwittingly clairvoyant declaration made at the start year by Choudhary is panning out accurately, as he spoke about going from data to decision more intelligently.

(Dr. Soudip Ray Chowdhury, CEO, Eugenie.ai)

Eugenie.ai, part of the larger Fractal Analytics network, is an AI-backed anomaly detection platform for data at scale. Chowdhury as its CEO, was engaged in a riveting discussion with the executive vice- chairman of Fractal Analytics Srikanth Velamakanni in July 2018 on the biggest challenge enterprises face today. “We came to the conclusion that making decisions efficiently is still something major enterprises across domains struggle with. At a time when AI and data insights have advanced by leaps and bounds, why is this still a challenge,” asked Chowdhury.

This led to the genesis of Eugenie.ai – by building a tech stack to handle large volumes of data enterprises are known to generate, building a suite of AI solutions on top of that to extract meaningful data, and finally involving a human being for validate these decisions and close the loop. After several rounds of discussions with industry leaders and reams of research, Chowdhury realized a few trends were prominent across the board: “Today, every enterprise is collecting a lot of data about their business, customers, vendors, competitors. This is essentially creating an information overload; and all this data is being channeled through legacy systems, ancient versions of dashboards and the like, which are simply not designed to competitively manage the velocity and complexity of data. As a result, enterprises are compelled to make hypothesis-driven business decisions, which are riddled with cognitive biases.”

The startup follows a three-pronged strategy of connect, track and diagnose (also referred to as the Spot Explore Exploit mantra) to achieve the goal of helping enterprises go from data to decision efficiently. The gamut of decisions being taken by organisations vary from tactical to strategic to operational – while the first two can still rely on conventional methods of deduction and reasoning, with precision not being a harbinger of success, operational decisions require a different approach. “Strategic and tactical decisions need time and brainstorming, and have a long term delivery goal. But operational decisions are rapid. Every second, every minute or every hour, businesses are making these operational decisions. There is little room for error here. This is where we come in,” explains Chowdhury.

(A snapshot of the dashboard by Eugenie.ai)

Notably, the most crucial component for Eugenie.ai from the three-pronged strategy is the ability to diagnose. This is where ML and analytics plays a very evolved and prescriptive role. The assimilation of data takes place through a network of sensors and micro-sensors installed at strategic positions across a machine or workflow, but the data streaming in from these sensors is what holds the secret sauce to the magic Eugenie.ai is capable of. “Using ML, we can correctly diagnose the problems, detect anomalies and deliver insights on what the problem is, and why a machine is not performing as expected. We provide a prescriptive framework that will enable the management to take the necessary steps.”

For instance, he recalls a project with a prominent O&G company where Eugenie.ai monitored product losses at tank farms. This was after the company reported of significant loss in gasoline volumes stores in these tanks. Eugenie’s anomaly detection tool figured out the variances in the product volumes at the time of unloading. The losses were mapped to the technical and operational determinants. The technical determinants for the product losses were - evaporation, density variation, ambient & fluid temperature variation and rainwater ingress in floating roof tank. This brought transparency in operations by building digital twins of the storage tanks. A control panel helped operation teams to track, diagnose and act on any discrepancies, which were then highlighted on a real-time basis allowing prompt action. Data to action time reduced from days to hours.

Eugenie.ai works with a range of clients across sectors like finance, ecommerce, telecom, oil & gas; and considers itself an enabler to larger solution providers like Microsoft, Qualcomm, SAP and Intel to name a few. Chowdhury says, “The Industry 4.0 is a $1.6 trillion market and there are sufficient opportunities for everyone to thrive. Moreover, it’s not entirely accurate to compare the capabilities of larger corporations and startups as both have different value propositions. However, due to their nimble nature and reduced levels of compliance, I do think startups have a chance to bag more PoCs and projects in the future. The focus for smaller companies should be primarily on delivering superior technology solutions.”

With a team of passionate engineers, data scientists, designers, sales and marketing professionals, team Eugenie.ai is on a mission to make operations more reliable by transforming people, process, and assets in every enterprise to be more efficient using AI, engineering, and design. “Operational reliability is the first step towards sustainability (less carbon emission and carbon footprint), and this is how we aspire to create a planet-size impact with our technologies and products.”