Teaching Machines to Think About Journeys

Custom AI models trained on organizational knowledge represent a structural shift in how journey management programs can operate. The shift is not primarily about efficiency — though AI companions do accelerate certain tasks significantly. It is about continuity: the possibility of an institutional memory that does not depend on any single person's presence, availability, or decision to document what they know.

SojournGPT, the AI companion built around Sojourn's methods and frameworks, is a practical expression of this possibility. It is trained on the book's materials and updated as methods evolve. Teams working through discovery, opportunity definition, or solution development can query it directly — for examples, for framework clarifications, for alternative framings of a problem they are stuck on. This is the application of the principle explored earlier in the context of AI as journey knowledge repository, taken one step further: not just storing knowledge, but making it conversational.

What an AI Companion Changes

The most immediate change is accessibility. Journey management frameworks are learnable, but they require time to internalize. A product manager who is new to the practice may understand that they should describe their work in terms of customer pains and journey stages, but struggle to translate that understanding into the precise language that actually advances the conversation. An AI companion trained on the methodology can provide examples, suggest rephrasing, and explain why a particular formulation is more useful than another — in real time, at the point of need.

This is different from documentation. Documentation requires the reader to know which section is relevant to their current situation. A conversational AI companion can be queried in the language of the problem: "We have a cluster of insights about customers losing confidence during the trial period. How should we reframe this into a How Might We statement?" The response connects the specific situation to the general framework, which is the translation work that typically falls to the journey orchestrator.

"A good AI companion does not replace the orchestrator's judgment. It scales the orchestrator's vocabulary — making the shared language accessible to everyone who needs to use it."

What It Cannot Replace

The AI companion accelerates the aspects of journey management that are primarily cognitive: generating options, clarifying definitions, drafting frameworks, testing whether an insight is well-formed. It cannot replace the aspects that are primarily relational.

The trust-building conversations that make discovery possible — the meetings where an account manager gradually becomes willing to share honest concerns about a customer relationship — require human presence, patience, and the specific kind of reading that comes from watching someone's discomfort in real time. The political navigation that keeps a Big Solution moving forward when the team's executive sponsor is distracted — that requires the orchestrator's judgment about what to say to whom and when. The emotional work of maintaining a team's motivation through a long, ambiguous creative process — none of these are tasks an AI companion can perform.

The boundary is worth being precise about. AI excels at the work that is legible: if the task can be described clearly enough to be prompted, it can often be assisted or automated. Journey management's most consequential work tends to be illegible: navigating the specific political terrain of a specific organization, reading the specific people in a specific room, knowing when to push and when to wait.

Using It Well

In teaching contexts, the AI companion functions as a sparring partner: generating fictional scenarios, stress-testing opportunity statements, producing examples of different confidence-tier insights for the student to practice evaluating. This is particularly valuable in classroom settings where the instructor cannot generate unlimited tailored examples.

In practice contexts, it functions as a framework reference and ideation accelerator: helping practitioners who know the methodology well to move faster through the legible parts of the work so that their time is available for the illegible parts that require human judgment.

The integration of AI into journey management practice is not a replacement of the methodology — it is an extension of the methodology's reach. The core still requires the careful, patient, human work of showing up to organizations and helping them see their customers more clearly.