A few years ago, a client of KPMG, the professional services and accounting firm, asked the company to help them tackle the menace of online frauds – like fake websites selling fake Louis Vuitton handbags. The problem was that there wasn’t any technology that could simultaneously monitor all the websites selling Louis Vuitton branded handbags and then help take down the fraudulent sites in a short time. So, KPMG’s India team built a solution from scratch called the Digital Signals Insights Platform (DSIP) in 2017 to provide their clients some cyber serenity. The platform has so far detected around 25,000 threats for 70 of KPMG’s clients and has prevented over 15,000 frauds from occurring worldwide. The platform helps clients identify, assess and act on insights arrived at from a set of digital risks, says Akhilesh Tuteja, global head of cyber security consulting and India head of digital consulting at KPMG. “DSIP is an intelligent, always-on platform that continuously monitors upcoming threats arising from cyber, social, mobile stores, dark web and surface web. It also provides real-time custom dashboards, real-time actionable insights and threat categorisation to enable takedowns,” he says. Building the platform took the combined effort of a large team. Configurable bots, APIs and crawlers had to be programmed to monitor upcoming threats, negative sentiments, and to red-flag keywords. The platform was also designed to alert users to malicious traffic domains or brand infringement cases. Reverse image search algorithms were used to help identify logo or image infringements. “Under the hood, machine learning and an AI-powered engine perform constant threat categorisation to help clients keep track of potential frauds,” Tuteja says. The team, Tuteja says, used an uncommon combination of skills. “We brought in people who understand cybersecurity, data analytics, big data, AI & ML, and even behavioural patterns and psychology,” he says. The platform has widened in scope from its initial conception. The team had to constantly adapt to new digital threat scenarios. When they started work in 2017, the aim was to identify fake websites. Over time, they realised that a lot of the same threat actors or criminals were also stealing data from social media platforms. So they built that feature in. “Later, we had to incorporate dark web monitoring into the platform, when the criminals migrated there. Now, the platform is in its third generation, and has a wide array of potential use cases,” says Tuteja.