“What does the model not know that you do?” Divya Parekh puts that question to every team she works with, and what comes back tells her how much judgment survived the rollout. The prevailing assumption is that AI adoption is a procurement exercise: choose the platform, train the staff, count the hours returned. “Leaders assume the risk is falling behind on tools,” she says. “The deeper risk is quieter: people treat the model’s confidence as a substitute for their own judgment.

Divya Parekh established The DP Group in Raleigh, North Carolina, and Thinkers50 recognized her as one of the Top Global 50 Influential Coaches.

The Real Bottleneck Was Decision Clarity

The constraint she finds inside stalled organizations is rarely the software. More often, nobody has agreed on what would count as enough evidence to act. Her first request is not for a system map. Before discussing any tool, she asks participants to bring three recent high-stakes decisions to the table, real ones with names and dates. The team records who decided, duration, plus where discussion halted. Then she encourages the team to test each decision against one question: “When you finally decided, what were you actually missing? Was it information? Or was it the confidence to act on information you already had?”

The pattern often unsettles the team. “Usually it wasn’t missing data. The team already had plenty of information. What they lacked was agreement on what counted as enough evidence to move,” she says. That diagnosis sets her apart from the two starting points CEOs most often encounter. An advisor who comes mostly from technology will often start with the platform. An advisor who comes mostly from coaching will often start with mindset. Her sequence puts neither first. “I start with the decision itself, and the right tools and the right mindset both follow from what that decision needs.”

That discipline traces back to the same instinct, honed over fifteen years in technical and process leadership roles in the biopharmaceutical industry, where Parekh helped advance products through complex, highly regulated environments. Lean Six Sigma was a core working discipline there. “You don’t move a product forward on enthusiasm. You move it through stage-gates, using Lean Six Sigma and core scientific disciplines, because a wrong step downstream is expensive and sometimes dangerous.”

Where the Agency’s Backlog Was Building

A PR and marketing agency came to Parekh with what appeared to be a capacity problem. Every client deliverable, including briefs, drafts, and reports, passed through senior team members for manual review, and the backlog limited how many clients the firm could take on. The obvious answer seemed to be another tool. Parekh first mapped how the work moved and where it stalled. She then built AI-assisted workflows around the repeatable tasks, incorporating the agency’s voice and standards and defining what a human reviewer would still examine before anything reached a client. Parekh reports that the redesigned workflow cut turnaround time by about 45 percent. “That didn’t put anyone out of a job. It freed the team to take on more clients with the same headcount, which was the actual goal.”

What surprised her sat on the human side of the room. “The team worried that faster would mean sloppier and that the work would stop feeling like theirs.” Reassurance alone did not resolve that concern. “The turning point came once they saw the human review step working and the quality remaining intact.” Looking back, Parekh would move faster on that conversation next time: “If I ran it again, I’d have that conversation on day one instead of letting it surface on its own.”

For executives who fear that AI will cost the organization its character, Parekh places the responsibility on leadership choices. “AI doesn’t decide how human your company is. Your people do, in the moments they choose to review the output, question it, and take responsibility for what the machine hands them.”