A walkthrough of every configuration tab in the BDT Design Studio — shown here using the live Afghan Women's Leadership scenario as it exists in the authoring environment.
My Future, My Choice — a scenario-based learning module designed for young Afghan women navigating education under Taliban restrictions. Players face consequential choices across a pre-generated decision tree with 5 levels of depth and 4 spiral cycles.
The scenario was built in this studio using Co-Authoring mode — an AI-assisted seven-step intake followed by a split-screen chat/tree authoring flow.
| Authoring mode | Co-Authoring (LLM + SME) |
| Branching mode | Pre-Generated |
| Tree depth | 5 levels |
| Spiral cycles | 4 |
| AI debrief | Off |
| Feedback signal | Implicit (narrative only) |
| Cascading consequences | On |
| Path reveal | On |
| Scenario seeds | 3 seeds provided |
Click any thumbnail to enlarge. Click a tab name below to jump directly to it.
Click to enlarge
The primary workspace. The main panel displays the full decision tree as a visual node graph — nodes connected by consequence paths, with color-coded quality flags on each node (red = review required, amber = review recommended, green = approved). Layout can switch between top-down, left-right, and compact tree views.
Click to enlarge
The foundational intake form. Captures the project title, course domain, level, and organizational context. These fields seed the AI's understanding of the scenario space before any tree generation begins. The organizational context field is a rich textarea that describes the setting in which the scenario will be used.
Click to enlarge
Up to three learning objectives can be defined, each with a priority level and Bloom's Taxonomy cognitive level (Recall through Evaluation). The AI uses these objectives to score and select activity types and to calibrate the difficulty and depth of each decision node. The Afghan Women scenario defines three objectives at Application and Analysis level.
Click to enlarge
Defines the learner profile in detail — not just role level, but psychological and contextual characteristics that shape how the AI calibrates the scenario's tone, complexity, and pacing. The Afghan Women scenario is configured for Individual Contributors with moderate cognitive load tolerance, trauma-informed pacing, and low prior game-based learning experience.
Click to enlarge
Controls the architectural shape of the game session. Branching mode (Pre-Generated vs Dynamic) determines whether the tree is fully authored before play or generated in real time during the session. Branching structure (Tree, Diamond, Network) controls the graph topology and convergence behavior. The Afghan Women scenario uses a pre-generated Tree structure with 5 depth levels and 4 spiral cycles.
Click to enlarge
A toggle panel for each active game mechanic. The Afghan Women scenario uses implicit narrative feedback (no explicit scoring shown to the learner), checkpoint-return replay, cascading consequences, and path reveal. The "Require Hard Choice" setting controls what proportion of nodes present genuine dilemmas with no obviously correct answer — set to roughly half for this scenario.
Click to enlarge
Controls the visual presentation of the player experience — not the studio UI, but what learners see when they play the scenario. Includes a colour palette editor and a border style library with 20 styles across 5 families. The AI can suggest a theme based on the course identity and domain. The Afghan Women scenario uses warm earth tones: cream background, dark brown text, and aged-parchment border accents.
Click to enlarge
Scenario seeds are the SME's input to the AI Co-Authoring process — structured descriptions of challenging situations from which the AI generates decision nodes. Without seeds, the studio operates in Just-in-Time mode (full LLM generation from intake data alone). The Afghan Women scenario provides 3 seeds, each describing a real dilemma: device sharing, support-network risks, and navigating family pressure. Each seed captures naive and sophisticated responses, downstream consequences, and cascade opportunity.
Click to enlarge
Configures which model tier handles each phase of the authoring and runtime pipeline. Three tiers are defined: Thinking (complex reasoning, tree architecture, quality evaluation), Normal (standard generation tasks), and Flash (lightweight UI-facing tasks). Each tier can use the platform-managed key or a BYOK (bring your own key) configuration. The model catalog loads dynamically from the API server — shown here in a dev environment where the catalog is unavailable.
Click to enlarge
The Integration tab is designed to configure how the scenario integrates into a broader learning experience — LMS connection, SCORM packaging, Helix Portfolio tracking, and post-session reflection prompts. This tab is currently in development and will be added once the integration schema is finalised.
Click to enlarge
The Transfer tab will configure the on-the-job transfer support mechanisms that follow the scenario — structured follow-up prompts, manager briefings, and 30/60/90-day check-ins that are part of the Experiential Spiral's fifth phase. Transfer is what separates a learning event from a behavioral change program. This tab is currently awaiting the transfer schema definition.
Click to enlarge
The AI Co-Author is not a separate tab but a persistent left-panel chat interface that's always accessible from the Tree Editor. Once intake is complete, clicking Generate Tree initialises the LLM with the full Master System Prompt plus the scenario seeds, and a co-authoring conversation begins. The SME can request new nodes, revise consequence paths, and approve or flag individual nodes — all through natural language. The tree updates live on the right as each node is generated.
Open the BDT Design Studio to explore this scenario in its live authoring environment, or contact GamaVida to set up a new scenario for your organisation.