IBM launches 'Forward Deployed Units' pairing AI agents with specialist teams to speed enterprise AI deployments

IBM launched a new delivery model for enterprise AI on May 14, saying its consulting arm will use “Forward Deployed Units” that combine small specialist teams with AI agents to move projects into production faster. IBM Consulting said the units are built as pods rather than individual roles, with human specialists directing digital agents that handle coding, testing, evaluation and documentation. The company said the model is already being used with Riyadh Air, Nestlé, Heineken and Pearson. The announcement adds detail to how IBM is trying to tie consulting work more tightly to its AI and hybrid-cloud products. ### What exactly is IBM putting into the field? IBM Consulting described a Forward Deployed Unit, or FDU, as a pod of senior people supported by a “digital workforce” of specialized AI agents. In the company’s May 14 announcement, Mohamad Ali, senior vice president and head of IBM Consulting, said the units are designed for “hands-on execution” rather than strategy work alone. (newsroom.ibm.com) The May 14 post said the agents inside those pods handle coding, evaluation, testing and documentation under human direction. IBM said the structure is meant to place people “at the edges” while software agents do repeatable middle-layer work inside the delivery process. ### Why is IBM calling this a unit instead of a forward-deployed engineer? (newsroom.ibm.com) IBM said the market has been focusing on the job title “forward deployed engineer,” but argued a single person cannot solve the operational problems that slow enterprise AI rollouts. In its announcement, the company said it has long embedded senior technical staff, including fellows and distinguished engineers, in client work, and is now turning that approach into a repeatable model. (newsroom.ibm.com) Computerworld reported on May 15 that the forward-deployed engineer title is spreading across the industry as vendors and service providers hire people to help customers install and scale AI systems. The publication cited LinkedIn data showing forward-deployed engineering roles increased 42-fold between 2023 and 2025, compared with 13-fold growth for AI engineer jobs over the same period. (newsroom.ibm.com) ### What problem is IBM trying to solve for customers? IBM said enterprise AI spending and experimentation have accelerated, but deployment remains slow because operating models still depend on adding labor rather than coordinating software agents, governance and execution. The company said many organizations can define AI goals and test tools, but struggle to move from pilots to production. (computerworld.com) Mohamad Ali wrote that the underlying obstacle is not vision or technology but delivery structure. IBM said the units are intended to compress the path from idea to implementation by combining domain expertise, engineering and AI-assisted execution in one team. ### How big a productivity claim is IBM making? (newsroom.ibm.com) IBM said a six-person FDU can do the work of a 30-person team. The company tied that claim to the use of specialized agents inside each pod and said the economics improve as methods are reused across engagements. That claim comes as other companies are building similar deployment-focused teams. (newsroom.ibm.com) Computerworld reported that Google had 1,513 openings for forward-deployed engineers and that OpenAI had 31 openings tied to its newly launched Deployment Company. The same report said Microsoft partnered with Accenture in March on a forward-deployment effort. ### Which customers are already using the model? IBM named Riyadh Air, Nestlé, Heineken and Pearson as current users of the Forward Deployed Units. The company said those engagements are aimed at moving AI from isolated pilots into production at scale. Yahoo Finance, citing the rollout on May 14, reported that IBM framed the units as senior-led pods for on-site client delivery. (computerworld.com) That report also noted IBM introduced two managed AI services on IBM Cloud the same day: Red Hat AI Inference and OpenShift Virtualization Service. ### Where does IBM say the rollout goes next? (newsroom.ibm.com) IBM said on May 14 that it is deploying FDUs “at global scale” across Asia Pacific, Europe and the United States, and increasing the number of units it is putting into the field. The company did not disclose a target number of pods or a timetable for the expansion. (finance.yahoo.com) IBM’s next public marker is likely to come through additional client deployments or consulting updates tied to its AI services portfolio. As of the May 14 announcement, the named participants in the rollout were IBM Consulting and the four customer companies IBM listed: Riyadh Air, Nestlé, Heineken and Pearson. (newsroom.ibm.com)