Project Type
Independent UX research + systems-design case study
UX Research Case Study
Scaffold
Executive function support for cognitive overload.
An independent UX research and systems-design project exploring how AI-supported tools might help adults experiencing task paralysis, overwhelm, or executive dysfunction move from stuckness into a next small actionable step—without shame, surveillance, or coercive automation.
Project Snapshot
Timeline
4 months
Participants
7 long-form interviews conducted; 6 retained transcripts used for deep synthesis
Session Length
60–90 minute semi-structured interviews focused on lived experience and support needs
Research Outputs
Interview guide, synthesis, interaction model, usability findings, ethical AI constraints, high-fidelity prototype
Core Question
How might support systems reduce cognitive friction while preserving agency, consent, and emotional safety?
Project Scope
Independent UX research and systems-design project developed over 4 months.
Scaffold was created as an exploratory research project investigating how support systems might reduce cognitive friction while preserving user agency, reversibility, and informed consent.
The project began as a task-support concept and evolved through interviews, synthesis, prototyping, and usability walkthroughs into a consent-based cognitive scaffolding system.
Problem Space
The problem is not motivation.
Many productivity tools assume that people already have enough cognitive bandwidth to define the task, prioritize the work, initiate action, recover from interruption, and respond to reminders.
For users experiencing ADHD, autism, depression, trauma, chronic illness, burnout, or executive dysfunction, the hardest part is often not the task itself. It is translating intention into sequence while already overloaded.
Scaffold explores a different model: not optimizing output, but reducing the friction between wanting to act and knowing how to begin.
Operational Relevance
Why this problem matters beyond the prototype
Scaffold is presented as an exploratory research and systems-design project rather than a commercial product. The underlying problem, however, is operationally relevant: people need enough cognitive bandwidth to understand systems, adopt tools, complete training, trust interfaces, and act on information.
In customer-facing products, cognitive friction can reduce trust, conversion, and follow-through. In workplace systems, training materials, HR tools, and AI-supported workflows, overload can limit comprehension, retention, adoption, and engagement.
Scaffold explores how support systems might reduce cognitive friction while preserving autonomy, clarity, reversibility, and user control.
Research Foundation
Primary research focused on lived experience, not assumptions.
I conducted qualitative research with adults who experience executive function challenges related to neurodivergence, trauma, chronic illness, depression, or cognitive overload. Interviews explored what “being stuck” feels like, what triggers shutdown, what forms of support help or harm, and what participants wished others understood about their internal experience.
Participant Criteria
Adults with lived experience of ADHD, autism, executive dysfunction, burnout, chronic overwhelm, depression, fluctuating capacity, or difficulty with task initiation and sequencing.
Research Ethics
Participation was voluntary; participants could redirect or stop discussion; emotional distress was not escalated for insight extraction.
Interview Protocol
Semi-structured questions explored lived experience, support strategies, emotional impact, autonomy, and unmet needs.
Evaluation
Prototype walkthroughs tested whether the system reduced task-initiation friction, clarified next steps, and preserved user control.
Evidence of Process
Selected research and design artifacts
These artifacts show how the project moved from research planning to synthesis, usability findings, ethical interaction constraints, and prototype refinement.
Research
Interview Guide
Research objectives, participant criteria, interview structure, and ethical research boundaries.
Open PDF →Synthesis
Research Synthesis
Affinity clusters, recurring patterns, and design implications from participant interviews.
Open PDF →Evaluation
Usability Findings
Observed interaction patterns, usability friction points, and resulting iteration decisions.
Open PDF →Systems
Ethical AI Constraints
Consent, transparency, reversibility, bounded scope, and non-coercive interaction principles.
Open PDF →Research Insights
What I heard
Across interviews, participants described executive dysfunction less as a lack of motivation and more as a state of overwhelm, shutdown, or disconnection from sequencing. Several described being capable of doing the task, but unable to locate the first step, retrieve the strategy, or tolerate the decision load required to begin.
Tasks become overwhelming when they arrive as an undefined package.
Participants often knew the task mattered, but the effort required to break it down created its own barrier.
Shame intensifies paralysis.
Support that feels corrective, controlling, or productivity-focused can deepen shutdown rather than reduce it.
Capacity fluctuates by context.
Energy, sensory input, social demands, stress, and health conditions all shaped whether a task felt possible.
Users need help that preserves agency.
The strongest support was not forceful automation, but consent-based scaffolding: small steps, clear choices, and the ability to pause without penalty.
Synthesis Snapshot
Affinity mapping translated interview data into recurring behavioral and emotional patterns.
I grouped participant observations into thematic clusters related to shutdown, cognitive overload, shame, capacity, support systems, and unmet needs.
Support failed when it increased pressure.
Participants described reminders, accountability, or help as harmful when it felt urgent, corrective, or controlling.
Shutdown was tied to shame and identity.
Task paralysis often created self-judgment, fear of failure, and a sense of being unreliable or incapable.
Capacity was context-dependent.
Energy, sensory input, emotional load, health, and social demands shaped whether action felt possible.
What Research Changed
The project moved from automation-first support to consent-based cognitive scaffolding.
Initial Direction
Automation-first task assistant focused on helping users complete tasks more efficiently.
Research Finding
Participants were sensitive to pressure, ambiguity, escalation, hidden complexity, and systems that felt controlling.
Final Direction
A consent-based cognitive scaffolding system focused on reducing friction while preserving autonomy, reversibility, and emotional safety.
Research to Design
How findings shaped the product
| Research finding | Design response in Scaffold |
|---|---|
| Users can freeze when a task feels too large, ambiguous, or emotionally loaded. | Scaffold breaks tasks into the next smallest actionable step instead of presenting a full productivity plan. |
| Shame and self-judgment increase cognitive load. | The interface avoids streaks, punitive reminders, urgency loops, or language that frames delay as failure. |
| External “help” can feel controlling when it removes agency. | Scaffold asks permission before taking action and allows “yes,” “not now,” or “pause” responses. |
| Users need support that adapts to available capacity. | The system offers lightweight pathways for focus, overwhelm, task initiation, and transition support. |
| Participants valued sensory quiet, predictability, and time to re-regulate. | Scaffold prioritizes simple prompts, low visual noise, and reversible interactions over feature density. |
Composite Persona
Alex
Composite persona built from recurring patterns across participant interviews.
Alex is a highly capable adult who has spent much of their life being responsible, reliable, and self-sufficient.
Despite strong problem-solving ability, Alex experiences recurring cognitive shutdown—moments where starting even simple tasks becomes neurologically inaccessible.
Triggers
- Unexpected demands
- High cognitive load
- On-demand communication
- Unclear starting points
Thrives With
- Predictability over urgency
- Autonomy and consent
- Low cognitive load
- Validation of effort
“I’m capable, but I need a safe way to begin.”
Understanding the Breakdown
Task initiation distress cycle
When task initiation fails, emotional and practical consequences accumulate, reducing cognitive bandwidth and reinforcing cycles of overwhelm.
Framework
Intervention
Scaffold as a cognitive prosthetic
Scaffold functions as a cognitive prosthetic: an external support that extends a user’s ability to initiate action when internal sequencing resources are unavailable.
Instead of requiring the user to plan first, the system:
Suggests a single first step
Reduces ambiguity
Externalizes cognitive load
Supports regulation
System Model
Interrupting the cycle
A supportive intervention restores bandwidth by making the next action smaller, clearer, and easier to refuse.
System Model
Moment of freeze
User cannot generate an entry point alone.
Scaffold prompt
System offers a concrete, manageable first step.
Action becomes possible
Support reduces friction and restores momentum.
Evolution
How the project evolved
Scaffold evolved from an early concept focused primarily on automation and task assistance into a broader exploration of cognitive accessibility, reduced emotional escalation, and consent-based support systems.
Phase 01
Early Paper Prototype
Initial concepts explored lightweight task support, conversational prompting, and interruption recovery.
Phase 02
Research + Systems Framing
Qualitative interviews shifted the project away from productivity optimization and toward cognitive load reduction, emotional regulation support, and autonomy-preserving interaction.
Phase 03
Interaction Design Refinement
The prototype evolved toward calmer visual hierarchy, reduced ambiguity, bounded focus interactions, and consent-based prompts.
Phase 04
Scaffold
The final framing positioned the system as cognitive scaffolding rather than behavioral automation.
Prototype
Prototype evolution
The early paper prototype explored three core pathways: starting a task, protecting focus, and responding to overwhelm. The high-fidelity prototype refined those pathways into a calmer, more structured interface that could support users when cognitive bandwidth is limited.
Home
Entry point for reducing activation energy and clarifying next action.
DMV Next Step
A concrete example of turning an overwhelming task into a first actionable step.
Urgency Calibration
Helps users assess real urgency instead of reacting to internal pressure alone.
Make It Smaller
Breaks a task into a smaller, more psychologically reachable unit.
Focus Timer
Creates a bounded, finite container for attention during initiation.
Transition chimes provide gentle sensory cues at the beginning and end of a focus interval, supporting users who benefit from externalized transitions and reduced reliance on internal time tracking.
Completion Moment
A brief celebratory transition reinforces progress and gives closure to the task.
Ethical AI Design Constraints
Designed for support, not behavioral control
Scaffold is designed as a support system, not a productivity optimizer or autonomous agent. Its behavior is intentionally constrained to preserve user autonomy, minimize unintended influence, and reduce cognitive burden.
- No action is taken without explicit user approval; there are no silent automations.
- The system is closed by default and does not access external data or other users’ information without opt-in.
- User data is isolated and not pooled, shared, or used to build behavioral profiles.
- All actions and recommendations are transparent and inspectable.
- The system avoids psychological pressure, including urgency loops, streaks, or compliance tracking.
- The user can pause, exit, or disengage at any time without consequence.
These constraints prioritize clarity, consent, and reversibility over automation and optimization.
Interactive Prototype
Explore the working prototype
See how the system supports users from overwhelm to action in a high-fidelity flow.
Open Interactive Prototype →Next Product Questions
What I would build next
If developed beyond prototype stage, Scaffold would require additional participatory research with neurodivergent users, stronger accessibility testing, privacy review, occupational therapy consultation, and careful evaluation of how AI support behaves under stress conditions.
The next product questions would be: When does support become pressure? What forms of memory feel helpful rather than invasive? How can the system stay useful while remaining transparent, optional, and easy to refuse?
Limitations
Research limitations
This was an independent exploratory study with a small qualitative sample. Findings are not intended to generalize across all neurodivergent users. The goal was to identify recurring patterns, generate design constraints, and evaluate early interaction concepts for emotional safety, clarity, and task-initiation support.
Because this project was independent, production readiness would require collaboration with product, engineering, privacy, accessibility, and clinical or occupational-therapy stakeholders.
“The project evolved from asking how systems can make people more productive to asking how systems can reduce the cognitive friction between intention and action.”
Scaffold reframes executive dysfunction as a mismatch between human cognition and system expectations—not as a failure of motivation.
The design challenge is not to push users harder. It is to reduce friction, preserve agency, and make action possible when internal resources are already strained.
Reflection
Design must support how people actually function.