People face an overwhelming number of small daily choices (what to eat, watch, or do) across countless platforms. Existing tools offer lists or generic recommendations but fail to reduce mental effort or help groups reach consensus.
The average adult makes tens of thousands of decisions daily. Constant micro-choices lead to cognitive overload, frustration, and delay, turning a 10–20 minute search into decisional paralysis.
Group decisions stall when shared preferences stay buried beneath individual bias. One person often becomes the default leader while others react passively instead of collaborating.
Many users find recommendation engines repetitive or opaque. Without transparent reasoning, AI feels magical and suspicious rather than like an informed friend.
Reduce Cognitive Load
Transform overwhelming daily searches into confident decisions made in minutes, not by adding more options, but by curating a small, high-quality set.
Facilitate Effortless Group Sync
Instantly visualize where group preferences overlap so decisions feel fair and collaborative rather than negotiated or leader-dependent.
Build Trust through Transparency
Design a system that explains why a recommendation was made, making AI logic legible and credible instead of opaque.
Three interconnected features translate research into a product that reduces friction for individuals and groups without removing human control.
Group members quickly input mood, budget, and deal-breakers. The dashboard automatically highlights alignment and removes conflicts, surfacing overlap before debate begins.
Built for high-pressure moments, this mode uses a timer and a “This or That” swipe interface to force commitment within a self-defined time limit, turning stalled group chats into action.
Every recommendation includes a bulleted breakdown of taste overlap, friend ratings, and logic (e.g., “Matches group protein goal”), so AI feels like an informed friend, not a black box.
Three decision styles, and recurring paralysis under opaque recommendations.
The team conducted surveys (13+ responses) and four in-depth interviews to understand how people make everyday decisions alone and in groups. Research surfaced decision-making styles (Proactive Leader, Mediator, and Cautious Adopter) along with recurring stressors like decisional paralysis and distrust of opaque recommendations.
Five epics: shared preferences, explainable AI, and time-bounded group decisions.
Insights were synthesized into affinity diagrams and POV/HMW statements, then translated into five product epics and user stories, including shared preference alignment, explainable recommendations, and time-bounded group decisions.
Violet + Lime design system: from Crazy 8s to high-fidelity flows.
The team used Crazy 8s for rapid ideation, mapped decision paths into detailed user flows, and iterated from paper sketches to a high-fidelity mobile design system. Violet and Lime colors convey a technical yet casual vibe suited to everyday social decisions.
Heuristics surfaced a prominent “Start a new poll” CTA on home.
I contributed to heuristic evaluations and cognitive walkthroughs with first-time users. Findings led to critical improvements, including standardized button styling and a prominent “Start a new poll” CTA on the home screen, before the team moved into high-fidelity validation.
Across research and testing, our team learned that users don't need more options. They need high-quality curation and a fast way to surface shared alignment.
Trust in AI is conditional. Recommendations are only acceptable when the system feels invisible in the best way: narrowing choices confidently while explaining its reasoning clearly. My work on heuristic evaluation and cognitive walkthroughs reinforced that small UI consistency fixes (button hierarchy, primary CTAs, and legible “Why This” copy) materially change how trustworthy the product feels.
Problems Addressed
Project Vision Achieved
WhatNow shifts the experience from negotiation and uncertainty to confidence and relief. By delivering a small set of vetted choices and surfacing overlap, it eliminates the bottleneck of deciding, letting users spend less time scrolling and more time enjoying their chosen activity.
