Enterprise leaders have navigated the experimental phase of generative AI. Balkrishan “BK” Kalra, Genpact’s president and CEO, declared this phase over during the Newsweek “AI Impact Forum” webinar hosted by Dr. Ranjit Tinaikar. Kalra stated, “Time of proof of concepts is gone. Time of experiments is gone.” The focus now shifts to scalable use cases.
Scaling up AI involves addressing issues overlooked during initial proofs of concept. Data must be usable, processes consistent across units, and employees need tool fluency. An accountable party must oversee agent actions. Kalra emphasized that scaled use cases should enhance business performance through growth, lean operations, or improved cash conversion. However, achieving these outcomes becomes complex when AI exits controlled tests and integrates with everyday operations.
Tinaikar referenced a study by Genpact and HFS Research, which surveyed 2,002 executives. Only 6% of organizations qualified as capable debt re-mediators, meaning they successfully addressed, executed, and measured solutions to such challenges. Tinaikar inquired what senior management should prioritize before advancing into agentic operations.
Kalra highlighted technology debt, citing old systems and patched technologies as visible issues. Hidden beneath are data, process, and talent debts, which restrict AI utility. Agents reveal these deficits since they rely on usable data and company-specific context. Business processes add complexity as they may differ by region, business unit, or practice. A standard AI model struggles to manage these variations without specific enterprise data and process insights.
Kalra illustrated how global operations require jurisdiction-specific handling. Much of the needed knowledge is enterprise-internal, often undocumented, demonstrated by employee exceptions and transaction management methods. “There is no artificial intelligence, no gains from artificial intelligence, if it is not coupled with process intelligence,” Kalra asserted.
IT and governance should engage early in conversations. When asked how companies can build IT department confidence in agentic systems, Kalra stressed early involvement. “Bring in upfront your CIO or a CDO partner,” he suggested.
Security concerns must be explored and addressed alongside a responsible AI framework. As agents execute finance, supply chain, and other operations, the potential for financial or regulatory mistakes grows.
Kalra described agentic operations as evolving towards machine-processed and human-validated processes, with humans handling exceptions and maintaining accountability. Workforce readiness is critical; workers need AI tool exposure to redesign work processes effectively. At Genpact, thousands have access to these tools.
Tinaikar connected workforce development to skill acquisition time. Training traditionally competes with daily duties. “It is not a privilege, it is an imperative,” he noted.
Kalra outlined two skill categories Genpact aims to develop. AI builders possess technical skills and relevant business domain knowledge. AI practitioners, initially skilled in finance, supply chain, or similar fields, must gain sufficient AI and data proficiency to engage with technology directly.
Kalra noted tasks and roles will significantly change as machines handle more executions. Tinaikar paralleled predictions of displacement during the smartphone era, which ultimately prompted new businesses and jobs.
Kalra agreed, noting that new technology brings new models and roles. He referenced Jevons paradox, where increased efficiency boosts overall use, leading to additional demand and activities.
Kalra cautioned workers, “Your job will not be taken by AI, but your job can be taken by somebody who knows AI better.” Improving model capabilities enhances automation opportunities, but implementing them necessitates usable data, defined processes, security measures, and adaptable employees.
Kalra summarized the issue, saying, “Aspirations are really high. Readiness is low.”
Leave a Reply