While artificial intelligence has enormous potential, there are still too many proof of concept projects and pilots. Most deployments target modest incremental improvement objectives.
In many organizations, work practices, incentives, and culture are still built around existing ways of working. Incremental improvements—like automating a report, improving customer service response time, or optimizing a workflow—have measurable before/after metrics and investments that can be justified more readily. Also, incremental AI fits into current leadership paradigms; while more ambitious AI deployments typically require redesigning roles, workflows, and decision rights. That’s much harder to do.
To achieve more ambitious goals with AI it’s important to recognize that transformative initiatives cut across silos and therefore demand a different way of viewing business. When goals are more ambitious, it’s essential to view how value is created for customers instead of looking solely at what departments do. That’s where process management enters. Process management expands AI ambition by shifting focus. Instead of a focus on small tasks and individual pain points, there is a shift to value creating workflows. Instead of fragmented data, there is a shift to looking at how data flows. Instead of just focusing on a tool, process management promotes a shift to thinking about the fundamental redesign of operating models.
Similarly, the likelihood of success with AI increases when companies simultaneously embrace knowledge management principles. That’s because knowledge management supplies institutional knowledge, domain expertise as well as how best to use tacit knowledge. Further, knowledge management creates the mechanisms to capture expert feedback, codify exceptions, and update rules and decision criteria.
Yet, few companies deliberately integrate process management and knowledge management practices when deploying AI. Process management (PM) and knowledge management (KM) typically reside in separate silos where process often sits in operations with an emphasis on Lean Six Sigma and knowledge management often rests in HR. Meanwhile, AI initiatives are regularly led by data/IT teams. Major opportunities are missed as these three initiatives are rarely integrated.
How can a process-based view of new product development (NPD) integrated with knowledge management principles aid the deployment of AI? A process-based view of NPD—especially when structured as a stage-gate system reinforces a process-based view and creates “decision points,” while KM ensures the right insights are available at each decision point. When integrated, they make AI far more deployable, reliable, and scalable.
Here’s a stage-by-stage summary for the first two stages in NPD.
In the Discovery stage a process focus (PM) helps identify opportunities and generate ideas while KM can capture market insights, past idea performance, customer feedback, and external signals.
Accordingly:
- AI models (trend analysis, idea generation) are trained on structured historical innovation data instead of scattered inputs
- Opportunity identification (e.g., unmet needs, emerging trends) is improved
- Reuse of prior ideation outcomes is enabled → avoids “reinventing the wheel”
In the Scoping stage PM helps with a preliminary assessment of ideas (market, technical feasibility), while KM curates benchmarks, prior feasibility studies, and competitive intelligence.
Accordingly:
- AI can rapidly compare new ideas against codified past cases
- Early-stage prediction models are enhanced (market size, risk scoring)
- Bias is reduced by grounding AI outputs in a broader knowledge base
The benefits of integrating PM and KM are equally compelling in the remaining NPD stages.
The bottom line is this: a process-based NPD structure tells you where decisions happen; knowledge management ensures what is known at those points; and AI amplifies how well decisions are made.
Integrating process management and knowledge management for AI isn’t just a technical task—it promotes thinking across structure, incentives, and how organizations get work done. That leads to a greater awareness of the importance of redesigning workflows, the importance of tacit knowledge, and how these two disciplines can work together to fundamentally redesign operating models. But it’s not easy and requires the following series of steps:
- Define and model the end-to-end process (not just isolated tasks)
- Layer in knowledge at each decision point (explicitly define the knowledge required)
- Apply AI to both process and knowledge at multiple points (not just one pain point)
- Design the new roles (don’t just keep old roles with new processes)
- Establish clear governance
In a nutshell - integrating PM and KM in deploying AI enables clarity around how work flows and how decisions are made—rather than just inserting AI into existing tasks. Is your organization ready?