This is the fourth in a five-part series on the software industry in the age of AI. 

No one knows exactly how AI will change the way companies look three or four years from now, but the contours are beginning to take shape. AI is fundamentally changing how software companies build, run, sell, and maintain software. The biggest impact will be felt by engineering and go-to-market organizations, where most of the headcount sits. One signal is that among leading software companies, revenue grew 22% faster than headcount in the past year, an indication of AI’s effect on employee productivity (see Figure 1).

Figure 1

Entire processes will be fundamentally redesigned in an AI organization, which will change roles, skill needs, and team structures significantly:

• Work will become more “end to end,” with jobs refactored around workers who stretch across what used to be separate roles. Product managers will prototype and even build. Engineers will talk directly with customers. Functional boundaries will blur, and teams will test and learn more as AI reduces the cost of trying.
• Teams’ structures will change, shifting from the standard “pizza team”—a team that can be fed by two large pizzas—to smaller units of three to five people augmented by AI agents. In engineering, this is already taking shape: The standard squad of one product manager and six to eight engineers across specialties is giving way to hybrid-agentic pods, which consist of a smaller number of engineers working alongside AI agents that handle coding, testing, review, and deployment (see Figure 2). The human activity shifts from writing code to supervising systems that write code.

Figure 2

• Individual contributors will start to look more like managers or orchestrators as they spend more time directing AI, dramatically increasing their productivity. Today, managers decide what work employees should do; increasingly, employees will decide what work AI will do. This doesn’t necessarily mean companies need more tenured, higher-titled people. It means individual contribution will shift from task execution to orchestration, strategic planning, critical reviews, and problem solving (see Figure 3).

Figure 3

• People who can think broadly at the high level but also go deep into detail—what the talent function calls “T-shaped”—will be in greater demand. AI can handle routine execution, so value will accrue to people with deep expertise and the ability to go wide.
• The pace will continue to quicken: Iteration cycles will shrink, speeding up innovation and product releases. Reddit says its teams can dream up an idea and have a functional prototype the next day. Customers will expect faster fixes and improvements and quicker returns on their software investments. To deliver at this pace, authority must sit closer to teams, and managers must shift from gatekeeping to enablement.

AI doesn’t respect org charts

As work and teams change because of AI, organizations’ structures will also need to adapt.

Reducing the number of layers in the organizational hierarchy (delayering) is already underway, and AI should accelerate this trend as it shrinks team sizes and speeds delivery. It should be easier for the front line to receive clear direction from strategic decision makers, with fewer layers between them.

As individual employees receive more autonomy to direct AI and make decisions, software companies risk organizational chaos unless their governance is built for an agentic world. Management systems must evolve, and decision rights need to be clearly defined: Where should humans stay in the loop, where can agents act autonomously, and who is accountable when things go wrong? As AI shifts decision making down the org chart, employees need clarity on where the company is headed and confidence that their judgment will be backed. Companies must shift toward an operating model that supports decentralized decision making.

Leading companies will also rethink their functional structures, not just making silos more porous but finding ways to architect AI-native processes end to end across formerly separate functions. A product-led company might tightly integrate in-product features, revenue marketing, and digital store operations. A large enterprise-sales company might couple account-based marketing, sales, and field delivery. The specific structures will vary, but the principle is consistent: You must be able to redesign workflows across functional boundaries, and the teams enabling that redesign need to be AI-first themselves—otherwise they become the constraint.

Change management is more critical than ever

Previous shifts to SaaS and Agile ways of working required new behaviors, but the adoption of AI could be even tougher. If every employee is managing a nonhuman agent workforce, they’ll need to learn to think like managers. Talent must stretch beyond current comfort zones. Embedding new ways of working can take years, and the risk of change fatigue is high.

Successful transformations treat workflow and workforce modernization as symbiotic efforts, continually adapting to feedback to amplify gains and correct imbalances by doing three things in parallel.

• Simplify and standardize the work: Implement zero-based redesign on end-to-end processes with AI embedded.
• Adapt the workforce: Define the roles and talent needed for the new processes and adapt the people as the work processes change.
• Manage the change: AI can’t succeed without people embracing new ways of working, and that requires visible support from the top. Senior executives need to guide teams through the change with clarity and empathy, defining a positive version of the future. The change should be reinforced throughout the organization, and there should be a closed-loop system of feedback, learning, and refinement driven by both human and machine intelligence. Training matters, but so does fast learning and open feedback. Above all, people need to feel supported as they take on new roles. Amid all this change, culture can be the most important of all. A culture that rewards experimentation and shared ownership will accelerate progress. One that clings to hierarchy and risk aversion will smother it.

There is no template for becoming an AI-first software company, but the destination is compelling: AI makes routine tasks easier and faster, freeing humans to focus on transformation. Software companies that succeed will become more adaptive, powered by innovation. Of course, this vision includes a daunting truth: Getting there will be messy, and there are no shortcuts.