Why enterprises are still struggling to get real ROI from AI

Despite the massive influx of funding into Artificial Intelligence (AI), many organizations find themselves entrenched in the experimental phase. From AI-driven assistants drafting emails to chatbots enhancing customer service, substantial operational changes remain out of reach. Malhar Kamdar, the Chief Growth Officer at Celonis, highlights a critical missing element: process intelligence. He asserts that AI has not fully realized its potential on the operational front due to its inability to grasp the intricate workings of businesses. "The business context required for AI to function effectively is dispersed across various systems, departments, and personnel. These operational blind spots hinder AI's ability to scale within enterprises," he explained during a discussion with Business Today. Celonis, a pioneer in process mining based in Germany, is shifting its focus from standalone AI applications to enhancing operational visibility within organizations. Their latest initiative revolves around a concept they refer to as a “context model,” which acts as a real-time digital representation of enterprise operations. This model enables AI systems to comprehend workflows, identify bottlenecks, and recognize dependencies throughout different departments and software. The company’s leaders argue that many enterprises rushed into adopting generative AI without first addressing the fragmentation in their workflows and outdated operational frameworks. Kamdar notes that while 95% of businesses are engaged with AI in some capacity, most of these efforts are limited to basic productivity enhancements—such as writing emails or summarizing documents. "True transformation in areas like supply chains and financial operations necessitates that AI understands the operational context," he stated. The widening gap between AI enthusiasm and tangible business benefits has raised concerns at the executive level. CIOs across various sectors are feeling the pressure to justify their AI investments in light of insufficient measurable returns. Dilipkumar Khandelwal, who has transitioned from CIO at Deutsche Bank to Chief Customer Officer at Celonis, observes a shift from mere AI trials to a focus on achieving returns from AI initiatives. "Many companies are investing in AI, but the pressing question is: how do you quantify the value created?" he remarked. He emphasized that process intelligence can provide a detailed overview of business operations, enabling organizations to pinpoint inefficiencies, automate workflows, and deploy AI agents effectively. Celonis boasts that its clients have realized over $25 billion in value creation through process intelligence implementations, with more than $10 billion of that being directly quantified by the clients themselves. Looking ahead, India is emerging as a key player in the enterprise AI landscape, bolstered by its expansive Global Capability Centers (GCC), digitally savvy startups, and a youthful workforce. Celonis has recognized India as a strategic market on par with North America and Europe, nearly doubling its workforce in the country over the past year. Kamdar believes that India's unique position allows it to bypass legacy challenges faced by Western companies, stating, "Indian firms are showing a much faster willingness to embrace AI-driven changes." However, even in India, organizations are realizing that AI cannot rectify flawed processes on its own. "The future of enterprises will be driven by AI and composable systems, but AI must function alongside existing workflows and human decision-making," Kamdar concluded. He emphasized that AI serves as a powerful enhancer for human capabilities, rather than a replacement.