Flow's key features in action (1 min watch)
Flow
assisting neurodivergent tech workers to optimize their workflow
Sponsor
Microsoft (Azure AI)
Timeline
6 Months (Feb - Aug 2025)
Team
Principal PM, Microsoft
Ainsley Tham (Product Designer)
+3 other designers
Context
Our sponsor, Microsoft, asked us to leverage LLMs and human-centered design to reduce workplace friction for professionals with ADHD. I designed Flow: a neurodivergent-friendly AI assistant and toolkit designed to minimize context switching, ease communication stress, and support sustained focus through features like Magic Mouse, Custom System Prompt, and Activity History.
Impact
Adoption & Usability
100%
all 4 usability testing participants stated that Flow would improve their workflow, and expressed interest in adopting it.
Time Saved
25 mins
with Flow, users context switch less and can stay in flow for longer, saving them a projected 25 mins/day.
Company Recognition
4+
Microsoft Designers and PMs acknowledged Flow's value and potential impact.
AI SYStem Prompt Satisfaction
4.3
the 25 surveyed participants rated their satisfaction with the AI system prompt an average of 4.3 out of 5.
What Microsoft Leaders Said
"This is such a real problem affecting many AI applications today. Love the focus on reducing cognitive load and it following you wherever you go! This is really cool to see!"
Senior PM @ Microsoft
"Wow, I'm looking at this and I feel like I want something like this. When I talk to my manager, this would be super helpful for me personally."
Senior Product Designer @ Microsoft
My Contribution
What I Did
System Prompt Design
Led the design, testing, and iteration of a custom system prompt, translating user research and SME insights into structured interaction rules that shaped the AI assistant’s behavior.
UI/UX Design
Designed and prototyped the Flow toolbar widget, from concept sketches to high-fidelity prototypes, focused on reducing context switching and cognitive load.
Research
Designed and executed mixed-methods research (surveys, semi-structured interviews, usability tests) across generative and evaluative phases, to inform both the system prompt and interface designs.
Overview
Flow: keeping neurodivergent tech workers in flow state
The Problem
Our research shows that neurodivergent people bring powerful strengths to their work: authenticity, hyperfocus and non-linear thinking. Yet, key pain points arise with the management of these strengths, disrupting their workflow.
Our Concept Key Features
Flow, is a non-intrusive toolbar containing features to reduce neurodivergent peoples' pain points and amplify their superpowers. Notwithstanding this, Flow is equally valuable to helping neurotypical people focus and sustain flow at work.
Magic Mouse
ADHD-Friendly System Prompt
Activity History
Key Feature 1: Magic Mouse
Magic Mouse reduces context switching and cognitive load by letting users capture, query, and act on information directly in their workflow: no copy-paste or app-switching required. It also dynamically shifts between "text highlight" or "area select" mode for adaptive, context-aware support.
Magic Mouse: Text Highlight mode
Where longer, multi-page text or precision is required.
Magic Mouse: Area Select mode
Where non-text assets (e.g. images) are involved.
Key Feature 2: Neurodivergent-Friendly System Prompt
I led the design of a custom system prompt of our AI Assistant to ensure neurodivergent-friendly communications with Flow. Through multiple rounds of user and subject-matter-expert testing, I iterated on the tone, response structures, and guardrails to keep interactions proactive, supportive, and low-friction.
Key Feature 3: Activity History
Activity History passively captures users work sessions through screen recording. It then creates a dynamic, searchable timeline that helps users quickly revisit and recover from interruptions, sustaining their flow and momentum.
Research
Pain Points of ADHD Workers' Workflows
Brief & Scope
Microsoft's Brief
Microsoft wanted us to leverage LLM to design an AI-powered solution for individuals with ADHD.
🔍 Scope: ADHD Tech Workers
Our target user group was Product Managers, Software Engineers and Designers; roles that face high-demand environments, that can amplify ADHD pain points.
Access
Microsoft’s sponsorship gave us direct access to tech workers for research and testing
Scalability
If our concept worked in this context, it has potential to scale to other knowledge workers
Business Impact
Aligns with Microsoft’s goal of supporting all employees and shaping productivity tools
Generative Research
Because ADHD is a sensitive, medical topic, we ran a mixed-methods study to understand pain points from both lived experience and scientific expertise.
Insight 1: Navigating Social Communication
Professionals with ADHD often feel emotional overwhelm or rejection sensitivity during social interactions. They need a judgment-free space to pause, rehearse, and regain clarity and communicate confidently.
Insight 2: Supporting Shifting Focus & Energy
Because ADHD causes fluctuating focus and energy levels, professionals need flexible systems that adapt to changing mental states, unlike rigid tools that can create additional friction.
Insight 3: Minimizing Disruptions & Context Loss
Professionals with ADHD experience fragile momentum and benefit from support that minimize disruptions, protect focus, and helps them quickly recover context so they can resume tasks without losing flow.
Problem
Untethering ADHD Professionals' Superpowers
Professionals with ADHD bring strengths like hyperfocus, non-linear thinking, and authenticity, but they also face recurring barriers in their workflows. Our research revealed that context switching, overwhelm, and workplace communication stress were the most consistent pain points.
The challenge: design a concept that amplifies neurodivergent strengths while reducing the cognitive and emotional friction that holds them back.
Design
Interface (UI) and Prompt (AI) Design
2-Pronged Approach
Designing Flow meant working at two levels: the interface layer, where users interact with tools like Magic Mouse and Activity History, and the conversational layer, where the AI assistant communicates through a custom system prompt. Both were informed by research and designed together to reduce context switching, cognitive load, and overwhelm.
Design Guidelines
Through the research process, I paid close attention to things that participants explicitly said that they did and did not want. I also synthesized overarching design guidelines that I wanted our concept to embody.
✅ Be quietly available, never intrusive
Create adaptive presence: showing up when needed, and invisible when not. Help should feel like a safety net, not a spotlight.
✅ Assist the user like you know them
Be proactive and align with users’ real-world patterns such as attention spans, habits, and energy cycles, instead of forcing rigid workflows.
❌ Don't create yet another ADHD App
In many cases, people have already found systems that work for them. Introducing yet another app would only increase cognitive load.
❌ Don't prescribe a one-size fits-all solution
Tools that work for one person might overwhelm another. Design for flexibility, adapting to different needs and coping strategies.
Ideation
We conducted multiple ideation sessions to brainstorm & refine concepts
Key Iterations
Toolbar Iterations
We designed a lightweight toolbar widget to reduce context switching and help users stay in flow. Research showed that professionals with ADHD experience context switching when support tools are scattered across apps. As such, we designed a consolidated hub where they could access the right help without breaking focus.
Feature Iterations
User testing and design critiques with Microsoft Product Designers revealed that less is more. By focusing on the features users valued most, we transformed the widget into a lightweight, always-available hub that supports ADHD workflows without adding friction.
The Final Design
Navigating Microsoft's Design System
Because Flow is intended to be part of the Microsoft ecosystem, we wanted to ensure congruence with Microsoft’s design language for familiarity and trust. That said, since we were designing a future-forward concept, we gave ourselves the leeway to push beyond system patterns.
SYSTEM PROMPT DESIGN
Customizing a Neurodivergent-Friendly AI Assistant
Why Necessary?
Our research revealed that 56.4% of participants struggle with workplace communication, often to the extent of avoiding responses when feeling emotionally overloaded. This is consistent with our Insight 1: "Navigating Social Communication", gleaned from our research.
😖 Over-Analyze
😪 Emotional Burden & Effort
Many participants already turn to AI as a judgment-free space for social scripting and practicing conversations. Instead of reinventing the wheel, we built part of our solution on AI, leveraging a tool users already trust to reduce emotional risk and support communication with confidence.
Our Process
🔨 Internal Evaluation
Each created a system prompt variation based on generative research, tested across the same 3 scenarios, and combined the best-performing elements into an improved prompt.
🔨🔨 ADHD User Evaluation
Ran semi-structured interviews where ADHD participants prompted the AI Assistant across 3 scenarios, sharing their expectations and real-time reactions.
🔨🔨🔨 External Evaluation
Used Azure AI Foundry to conduct a likert-scale evaluation, ran 25 user surveys, and conducted semi-structured interviews with a psychiatrist and an ADHD coach to ensure the responses were clinically appropriate and aligned with real-world practices.
How It Differs From Standard GPT
Comparing the response to the same input, our customGPT is proactive, supportive and low-friction. Designed for neurodivergent users, it prioritizes psychological safety, empathy and actionable steps.
Standard ChatGPT
✅ Our CustomGPT (Final Version)
Final System Prompt
I led the design of Flow's interaction blueprint, designed on LLM GPT-4o, deployed via Microsoft Azure AI Foundry. I designed it to be modular and formatted, instead of a single chunk of instructions, making it easy to make precise and targeted iterations post-testing.
Our system prompt was based on guiding principles that emerged from our research and testing
🏗️ Proactive Scaffolding
Anticipate moments of overwhelm, startup effort, and the “blank canvas” problem, and offer simple starts that lower effort and build momentum.
❤️🩹 Emotionally Safe
Reframe stress as situational, not personal, and designed responses that validate without clinical labels or judgment.
⚡️ Frictionless Output
Provide immediately usable and editable output to minimize effort and cognitive load.
📝 Task-Oriented (Action First)
Prioritized concrete and actionable outputs, and offer deeper reflection after.
🧭 Adaptive Role-Blending
Blend roles fluidly to keep support relevant to users' shifting needs.
Wrapping up
Handing Off to Microsoft
To hand off, we gave a final presentation to our sponsor and Master’s program professors. We also delivered design artifacts, including high-fidelity prototypes, the system prompt, and research documentation.
➡️ The plan moving forward
Our Microsoft sponsor told us that they will adapt and implement Flow’s designs, using our deliverables as a foundation for future iterations.
Product Demo (4 mins)
My team scripted, acted, filmed and edited a product demo showcasing Flow in action.