The Myth Of The 'AI Expert': Why The Fastest Path To Value Starts With The Team You Have

AI adoption in enterprises is no longer about novelty. Stakeholders demand results. Initial pilots dazzled teams, but a large portion of the initial enthusiasm has given way to intense ROI scrutiny as boardrooms worldwide are confronting the costly reality of AI pilots that fail to deliver measurable value. The gap between expectations and reality has also handicapped the 'all gas, no brakes' mentality often exhibited by enterprise AI adoption, as harder questions like "What's the business value?" and "What outcomes can we prove?" start to become the prevailing voices of reason.

We spoke with Prafulla Patil, the Founder and CEO of OneSolve, and former CIO/CISO at Mapbox, with leadership roles at LinkedIn, Atlassian, and Flexport. Having overseen enterprise AI and automation at multiple Fortune 500 firms, Patil knows where projects stall and how to get them unstuck. Patil argued that as the initial AI fatigue sets in from failed pilots, a structured playbook is emerging for enterprises ready to turn pilots into real results, starting with one simple principle: define your goal.

• Start with the goal: Patil said that every AI initiative should begin with the end in mind. "You can start backwards: define your goal and your KPIs, and exactly what you are trying to do." Rather than focusing on the latest tools or technologies, organizations should orient their efforts around measurable outcomes, ensuring that every AI investment drives real business value.

• Focus on the problem: According to Patil, successful AI projects start with understanding the problem, not the solution or tool. “Talk to your users as often as you can: shadow them, survey them, and ask what tasks they perform mechanically or repetitively,” he said. Leaders and frontline staff often see workflows differently, so gathering input from both groups ensures that automation addresses meaningful pain points rather than perceived priorities.