Ronny Fehling says stalled AI pilots are often misdiagnosed as technology, data or change-management problems.
“oftentimes this slowdown gets misdiagnosed. It is technology problem, it's a data problem, it's a change management problem.”
Person
Ronny Fehling
1 episodes · 18 quoted claims
Fehling says Anthropic, OpenAI, Blackstone and another firm each announced their own AI deployment company with private equity backing, totalling more than $5 billion of fresh capital.
“They all announced their own deployment company with the, you know, big private equity houses, roughly more than $5 billion of fresh capital”
Fehling recalls OpenAI saying earlier in the year that the limiting factor for enterprise AI value is not model intelligence.
“even OpenAI, I think, was saying earlier this year that they see that the limiting factor for seeing the value of A I in enterprises is not the model inter intelligence”
Fehling says Blackstone has described the scarcity of implementation engineers as one of the most significant bottlenecks to enterprise AI adoption.
“scarcity of these implementation engineers I is one of the most significant bottlenecks to the adoption of A I in enterprises”
Fehling worked with a large telecom operator to identify a bounded, AI-assisted configuration workflow as a first production slice.
“I worked on a large telecom operator and there we found, we identified a bounded A I assisted configuration workflow.”
In the telecom case, the team was asked to provide an ROI calculation for a program of first six and then twelve weeks.
“We had to put the R O I in there for first a six and then a twelve week program.”
Fehling says the telecom initiative slowed down because of added approvals, specifications and governance, leaving less room to iterate or fail.
“More approvals, more specification, more governance, less room for iteration, less room for failure.”
Fehling says that by the time a slowed-down AI program is noticed, twelve to eighteen months have often passed.
“And then by the time that anybody calls that out, twelve or eighteen months are gone.”
Fehling describes a manufacturing project in a highly regulated discrete-manufacturing environment.
“there I can tell we had a manufacturing case in discrete manufacturing, uh, quite tough, highly regulated.”
Fehling's team built a very small system that helps factory operators handle non-quality events faster, and it was adopted instantly.
“We built this very small system that helps these operators, blue collars, handle non quality events faster.”
Fehling describes a production gate where, if results miss expectations, the initiative is either killed or narrowed further.
“if the numbers don't match to where we think it should be, we kill it, or we have to further narrow it down.”
Fehling says an AI operating model often evolves from centralized to more distributed as it succeeds.
“It goes often from a centralized to a more distributed model.”
Fehling says a CEO told a steering committee that AI use cases should have a 'wow factor'.
“I had a C E O say, you know, that have this wow factor”
Fehling compares today's push for engineering foundations with the data-lake wave of about ten years ago, and pushes against it.
“We had this ten years ago around data lakes, and uh, that's also where I would push against”
Fehling cites a company that spent serious money on data modernization, hoping for product transparency and better margins.
“worked with a company that serious money into data modernization”
Fehling predicts a wave of top-down mandated AI rollouts into private-equity portfolio companies over the next twelve to eighteen months.
“I think we're going to see a wave of top down mandated A I rollouts into these portfolio companies over the next, you know, twelve, eighteen months.”
Fehling says a first AI production slice should be bounded enough to finish in six to twelve weeks.
“It needs to be bounded from the side enough so that we can finish something in six to twelve weeks.”
Fehling says leaders must prove AI works inside production, with real users, data and validation, before scaling it.
“you need to prove that this thing works in the inside production.”