Skip to content

Making AI Pay Its Way

How do we make artificial intelligence (AI) pay its way? One recent survey by Bain suggests that an overwhelming majority of enterprises are yet to find the right answer. Apparently, more than four in five company executives are still unsatisfied with how they are rolling out AI technology, and how they are gaining value from it. Given the relentless momentum behind investment in AI, machine learning (ML) and data science, that’s a frankly staggering statistic. But as Girish Agarwal, chief digital and information officer at Piab, explained in this Future Says interview, the key lies in recognizing that this is as much a question of culture and behavior as it is technology.

Agarwal is well placed to share wisdom on drawing value from AI. As a PhD fellow at the KTH Royal Institute of Technology in Stockholm, Sweden, he is researching how AI can transform value perception with customers and disrupt existing business models. His current employer Piab specializes in automated process components for gripping, lifting, and moving applications. The fact that the company name was inspired by the founder’s love of the mathematical formula Pi would suggest it’s fertile ground for AI and data-driven business transformation.

Building on many of the ideas raised by Aiko Yamashita in our first Future Says interview, Agarwal emphasizes the need to forge the right relationships between people, data, and technology. Enterprises and organizations should think in terms of how they can best convince their employees (and skeptics) of the veracity of a radically different approach. Invariably, realizing a data-driven revolution involves not only new technologies, but also innovative business models for capturing and delivering value.

For Agarwal, actions speak louder than words. Any AI center of excellence should be motivated by a desire to create “end-to-end prototypes, not PowerPoints.” Determination to go beyond proof of concept is another important mindset. To win buy-in from their colleagues, a data science team also needs to demonstrate tangible returns from new applications. In this respect, relatively modest, incremental gains are far more powerful than overly ambitious, blue-sky projects that fail to deliver on their promises. In the commercial domain, Agarwal believes that the ideal AI developments are measured in weeks, rather than months or years.

Just like Yamashita, he has some interesting thoughts about what makes a good data scientist. Agarwal certainly considers it another important question for enterprises to address. Investment in a data science center of excellence will invariably be needed to provide the foundations of any transformation. Ultimately, though, data science and insights from data-driven technologies should be embedded throughout the business. In achieving these goals, enterprises should not expect to find a large pool of talent available in the wider job market. Instead, the emphasis should be on building that talent for themselves and embracing the principles of democratization. The age of the citizen data scientist has well and truly arrived. 

Agarwal is clear that there are no magic bullets. For every organization, the transformation will be a journey, not a quick fix. But he is equally convinced of the need for urgency. “There is no right time to start.” Instead of looking for that perfect moment, the priority is simply to get moving. None of the apparent obstacles to change will be insurmountable. Ultimately, the most important step is simply the one that begins the migration to a data-led approach.

If Agarwal’s perspective resonates, then both his interview and Altair’s on-going Future Says series has much to offer. Series two of Future Says kicks off in September. Please register here for updates.