Work rarely happens in a single tool. Handling a customer issue might involve reading an email, checking an account, consulting a procedure, asking a colleague for context, and recording a decision. The skill lies in bringing those activities together.
For instructional designers, that creates a challenge: how do you give learners an opportunity to practice the whole task, with the information, tools, and decisions it requires?
SceneSnap scenarios provide a simulated desktop where documents, emails, AI stakeholders, and activities form an evolving work situation. Learners investigate, communicate, produce work, and make decisions within that environment.
AI-built interfaces extend what that environment can contain. Authors can describe a software interface in natural language, test and refine the generated result, and integrate it into a scenario. A simulated support workspace, opportunity record, or administration console can become part of the work learners practice.
The value is the ability to design a connected work experience around a learning objective—including the software through which learners need to act.
Give learners somewhere to apply their knowledge
Knowing a procedure and applying it to a messy situation are different demands.
A learner may understand an escalation policy but still need practice recognizing when it applies, finding relevant information, handling pressure, and documenting a decision. An exercise can bring those demands together when it provides the environment in which the work happens.
In a SceneSnap scenario, an email might introduce a customer issue. A document might explain the approved process. An AI colleague might offer additional context. A custom interface can provide the case record the learner needs to inspect and somewhere to submit the next step.
Each element serves the same task. The learner must connect what they read, what they hear, and what they do.
For an instructional designer, this opens up practice focused on how work is performed: selecting relevant information, managing uncertainty, applying a rule, and following through on a decision.
Shape the software around the learning objective
The software needed for an exercise depends on the role and the task.
A sales scenario might require an opportunity record with account history and discovery notes. A service scenario might need a ticket queue and an escalation form. An operations exercise might require a workspace for reviewing requests.
With AI-built interfaces, the author describes the screens, information, controls, and behavior the exercise requires. An AI agent generates the interface, and the author can try it and request changes through further natural-language instructions.
This approach is often called vibe coding. For instructional designers, its practical value is the ability to shape a working interface through a conversational authoring process.
The starting point should remain the learning objective. What must the learner inspect? Which decisions should they make? What actions must be possible?
A focused interface can provide the functionality needed for that practice. Recreating every feature of a business application may add complexity without contributing to the objective.
The generated result still needs review. Authors should check that its information is consistent, its controls behave as intended, and its design supports the task.
Make the interface part of an evolving situation
A custom interface sits within the broader scenario desktop, alongside the other resources and activities the learner encounters.
The scenario can supply case-specific data to the interface. Meaningful actions that the interface is designed to report can also become milestones that prompt the scenario to progress.
This makes it possible to connect software interaction with the surrounding work situation. A learner might review a request, consult a stakeholder, submit a decision, and then encounter a development that requires further action.
Those connections need deliberate instructional design. Submitting a form establishes that the learner submitted something. Whether the submission was appropriate depends on the situation, the learning objectives, and the evidence available for evaluation.
What could this look like in practice?
The following examples are proposed learning designs, rather than ready-made templates or reported customer outcomes.
Customer support: investigate and resolve an escalation
A learner receives a complaint about an unresolved issue. The simulated desktop contains the customer’s email, a policy document, and a support interface with fictional tickets and interaction history.
The learner investigates the record, speaks with an AI stakeholder, and decides how to proceed. They then document the decision in the interface and explain the next step.
The exercise can bring together information gathering, policy application, communication, and record keeping. Its learning objectives can address both the decision and how the learner reached it.
Supplier onboarding: apply a procedure under pressure
An urgent request arrives to onboard a new supplier. A custom interface contains the supplier record and submitted information, while a document explains the required checks.
The learner discovers missing information. An AI stakeholder presses for progress, creating a reason to clarify what can proceed and what needs further review.
The task is to investigate, apply the procedure, record a decision, and communicate a defensible next step. The interface gives the learner somewhere to carry out the operational part of that work.
Sales: turn discovery into an accurate next step
A learner prepares for a conversation using a simulated account and opportunity record. During the conversation with an AI prospect, they explore needs and clarify assumptions.
They then update the record and document an appropriate next step.
This connects the conversation with the work that follows it. The learning design can examine whether the learner distinguishes confirmed information from assumptions and whether the recorded next step reflects what was discussed.
Build evidence around meaningful actions
A complete work episode can reveal more than whether a learner knows the right answer. It can show how they use information and follow through on decisions.
Custom interfaces contribute the meaningful actions they are designed to report to the scenario’s interaction history, alongside other activity from the exercise.
That gives instructional designers another source of evidence to consider when designing evaluation and feedback. It does not automatically capture or assess every click.
Authors need to decide which actions matter and check that the interface communicates them. They also need to distinguish task completion from performance quality. A submitted escalation, for example, may be evidence that an activity was completed; assessing its appropriateness requires context.
Reuse the environment and vary the challenge
An interface is a reusable asset. The same support workspace can serve different scenarios, with case-specific data providing different situations for learners to handle.
This allows the environment to remain familiar while the challenge changes: a routine request, an incomplete record, or a case requiring escalation.
Interfaces also have saved versions. A configured scenario retains its selected version, and an author can choose when to adopt a newer one.
For instructional designers, that provides control over iteration. An interface can be refined and reviewed before an existing exercise adopts the change.
Keep the experience grounded in simulated work
These interfaces operate within the simulation. They do not connect to the organization’s live CRM, ticketing system, or administration console.
A submitted request or updated record represents an action in the exercise. Authors can use fictional data and build the functionality required for practice within that boundary.
The full experience still needs testing: the interface, the surrounding scenario, the actions reported, and the resulting feedback should work together toward the intended learning objectives.
AI-built interfaces give instructional designers more control over the environment in which practice happens. Combined with SceneSnap’s simulated desktop, they make it possible to design connected work episodes in which learners investigate, communicate, decide, and act.
Start with the work someone needs to perform. Then build the situation—and the tools—that let them practice it.
