As enterprises move from AI pilots to real-world deployment, IBM is rethinking both how its consulting teams are structured and how it addresses the structural gaps holding back large-scale adoption—shifting focus from experimentation to delivering measurable business outcomes.
The company is increasingly deploying what it describes as “forward-deployed” teams—integrated units that sit close to the client and combine engineers, architects, user experience specialists, data experts and industry practitioners—moving away from traditional siloed models.
Multi-Disciplinary Squads
“Now with AI, this whole concept of forward-deployed engineers is becoming real because these engineers have to live, breathe, speak, and hear the client’s lingo even more so (now) than ever before,” Juhi McClelland, Managing Partner, IBM Consulting, IBM Asia Pacific, told FE.
She added that these teams are being restructured into cohesive, multidisciplinary units. “We are now creating forward-deployed squads that have engineers, architects and an industry expert. These forward-deployed units are the ones making AI real for clients,” McClelland said.
This shift reflects the nature of AI deployments, which require alignment with business processes, domain language and organisational context, as well as a stronger focus on user experience to drive internal adoption.
At the same time, IBM sees enterprises struggling to scale AI beyond proof-of-concept stages. “In the last two years, there’s been a lot of experimentation. Of nearly 150 POCs only 25% have given an ROI,” McClelland said.
Overcoming the Four ‘Debts’
The company attributes this gap to four structural constraints that continue to limit enterprise-wide adoption.
“We found that there are these four ‘debts’ that exist that are preventing enterprises from adopting AI,” McClelland said. “One is the process debt. For the last many years, even when people were deploying AI, they were doing patchwork and not dealing with the root cause.”
She further identified technical debt, data fragmentation and talent gaps as key barriers. “Then there is technical debt, data debt (data sitting in puddles instead of an integrated, common view), and skills debt—not only skills to use AI, but also, as we strongly believe in ‘human plus digital,’ the people to use it,” McClelland said.
As enterprises push for clearer returns, the conversation is also shifting towards accountability. She said businesses are realising that to see returns on AI investments, they need to build the business case into their balance sheets and strategy.
With AI investments increasingly tied to financial outcomes, the focus is moving away from experimentation towards demonstrable business value. “Honestly, value realisation is the biggest conversation I have with every client,” she said.