The Adoption Gap
Two decades from freelance developer to enterprise AI builder and chief executive, Alkhafaji argues that AI has turned services from a people business into an outcomes business, and that adoption is the only real bottleneck.
"Don't settle. Always challenge yourself."
Dr. Ali Alkhafaji is Chief Executive Officer of Apply Digital, a business of around 700 people that he positions as an agentic customer experience partner, delivering content, marketing, data and commerce work through agentic AI workflows and delivery teams that pair humans with agents. He is known for arguing that AI has changed the economics of services from billable bodies to fixed price outcomes, and that adoption, governance and training decide who gets the value.
Alkhafaji started out as a freelance developer, co-founded Distiller, and built and sold TA. Alongside that he took a doctorate in video game psychology and educational theory, which he credits with teaching him to think critically about every conversation and with building the emotional intelligence he now screens for in hiring. At Omnicom Group he built and deployed Omni AI, the group's agentic platform, and describes seeing it land in the hands of a 100,000 person organisation as promising but incomplete, because the harder problem was persuading people to use it.
He joined Apply Digital as Chief Executive Officer wanting to take that class of solution to every client rather than one group. There he has built a custom internal training programme with role specific tracks and a badging system, using content from partners including Anthropic and Google, standardised on Google Cloud, which he describes as the only cloud platform spanning models, agents, cloud and data, and set the business on a path to ISO 42001, which he describes as the first certification focused on AI governance, safety and security. Alkhafaji says he rarely stays anywhere beyond two or three years before boredom sets in, and that leading a small company into a segment with no blueprint is both satisfying and terrifying.
AI will not replace your job, someone with AI will replace your job
"The separation between good companies that do this well and, and not is training, governance, and enablement."
Alkhafaji rejects the idea that the leaders are pulling away on tooling, pointing out that everybody has access to the same models and can now afford them. What divides the field, in his account, is whether a company trains its staff, enables them with the latest technology and wraps governance around the result so the brand is not exposed. At Apply he has built the framework in house, with role specific tracks and a badging system, and treats ISO 42001 as a teaching process rather than a badge.
"So you always have to have a human in the middle to be able to authenticate and authorize these things from being published."
For all his enthusiasm, Alkhafaji is unsentimental about what the models are: unpredictable, indeterministic and probabilistic, capable of producing output that is wrong, unsafe and risky for a brand. That is why he insists a human sits in the middle to authenticate and authorise anything before it goes out. Agents, in his view, earn their keep in concepting, ideation and scaling content, with the judgement kept human.
"We need to make sure that both sides are incentivized to innovate"
Alkhafaji argues that outcome based, fixed price engagements only work when the agency and the client both have something to win. In his model the client takes the AI discount, the agency protects its margin, the staff get the tools and the outcome is what everyone is paid against. Where the benefit flows one way only, he says the relationship is not a healthy one.
Stay close to the conversation.
New leaders and insights. No noise.