IN-PROGRESS
Helping Event Coordinators Manage Live Chaos with AI Voice
Research phase complete and the team has now moved into the design phase. This case study covers the research journey; design exploration is in progress.
ROLE
Product Designer
TIMELINE · TEAM
On-going · 4 Interdisciplinary Product Designers
MY CONTRIBUTIONS
Desk Research, Interview Moderation, Competitor Analysis, Synthesis.
THE PROBLEM
Event coordinators experience cognitive overload when unstable information, time pressure, social expectations, uneven availability, and environmental constraints make it difficult to keep people, decisions, and resources aligned in real time.
THE DESIGN CHALLENGE
Design a context-aware AI voice communication tool that supports coordination during live events, where coordinators filter, adapt, route, and track critical information across people, channels, and event environments, to reduce their cognitive overload.
MY DESIGN THINKING
RESEARCH
I argued against two of our candidate groups: drivers and dispatchers. Positioning the tool as 'helpful' for drivers could just as easily read as encouraging multitasking behind the wheel, already a high-attention task on its own. I'd proposed dispatchers myself, hoping to ease their anxiety, depression, stress, and PTSD, but reversed my own position: dispatching is too high-stakes for AI, it affects human lives directly and already runs under regulated safety protocols.
COMPETITOR ANALYSIS
I contributed to review 6 existing tools for voice communication and those were strong in exactly two places: before the event and after it. None of them touched the live moment, where the actual chaos happens and that's the opportunity.
SYNTHESIS
Pattern-matching pain points is the easy version of synthesis. The harder version is refusing to stop at the first pattern that satisfies the room. I pushed back on my own team three times before we landed on cognitive overload instead of “the tool was bad.”
01 — SITUATION
In January 2026, our team was challenged to imagine the future of AI-powered voice communication between businesses and people. Because AI would become part of these conversations, I saw trust as a foundational design question: what role should AI play, and what would make people comfortable relying on it? I had already integrated AI into my daily work as a thinking partner, but I repeatedly encountered sycophancy and hallucinations. These behaviors made AI feel unreliable and showed how quickly trust can break down when its responses are inaccurate or overly agreeable.
To understand what trustworthy AI behavior should look like, I examined how trust develops between humans and AI agents. The research suggested that trust grows when AI behaves cooperatively, responds with social awareness, and supports the people involved rather than acting separately from them.
This led to a clear design principle for the voice channel: AI should strengthen the interaction between people, not replace either person in the conversation. That principle became the lens I used to evaluate every concept that followed.
This case study focuses on the decisions where my research, judgment, and advocacy shaped the team’s direction. From the evidence I introduced to the trade-offs, methods, and synthesis decisions I pushed forward.
02 — RESEARCH
Our preliminary research question started broad: how do hands-busy people use voice interaction technology to navigate personal and professional needs?. “Hands-busy” was too generic to design against, so we scoped further and began researching other hands-busy populations, defining them as “multitaskers” and building out metrics for what that meant: multitasking during calls, how they prioritize, and how their attention splits.
I started this scoping work by identifying beneficiaries: people whose calls are tied to money, safety, scheduling, or follow-through including:
Remote and hybrid workers
Medical professionals (physicians and APPs in tele health and hybrid practice also hospital and inpatient unit coordinators)
Emergency dispatchers
On dispatchers specifically, I found research showing their cognitive load increases their risk of anxiety, depression, stress, PTSD, and other physical health issues. For remote and hybrid workers, my read was that humans are cognitively limited, but work environments demand constant multitasking. The problem is environmental, not behavioral.
The Team's High-stakes & High-frequency-multitasking Mapping Used to Prioritize Target Users.
As a team, we then organized these candidates along two axes: high stakes and high frequency of multitasking. We defined “high-stakes” as occupations involving human safety, health, or social consequences. That exercise produced three groups of target users to research further:
People who orchestrate others: event planners, wedding planners, fundraising officers, real estate agents, sales, contractors, recruiters/talent acquisition
Public safety tele-communicators
I laid out the trade-offs for the team and argued against two of our three candidate groups, against driver and public safety tele-communicators. First, if we chose drivers (truck, ride-share, chauffeurs) as our target users, I flagged that this could read as a double-edged sword. A tool positioned as “helpful” could just as easily be seen as encouraging multitasking during driving, when driving itself already demands high attention.
Second, I had actually proposed dispatchers earlier myself, motivated by a genuine wish to reduce dispatchers' risk of anxiety, depression, stress, and PTSD. But after thinking it through more carefully from a different angle, I reversed my own position: dispatching is too high-stakes for AI to be applied here. It directly affects human lives and operates within strict agency protocols, professional standards, medical oversight, and legal requirements.
THE RESULT
The team agreed with both arguments I pushed on, and we selected “people who orchestrate others” (event planners, personal assistants, and similar roles) as our direction.
With a direction chosen, our research question sharpened:
How do event coordinators use voice communication to keep events running smoothly while managing emerging needs in real time?
We framed event coordinators as “orchestrators of people”: they manage vendors, clients, venues, and teams simultaneously across overlapping calls and competing priorities. Yet little was understood about how they actually manage attention, capture information, and ensure follow-through.
COMPETITOR ANALYSES
I contributed directly to 6 out of 16 competitor analyses and I learned that most tools today are strong in two phases: pre-event planning and post-event analysis (recording, transcribing, and extracting or generating action items). But both of these are retrospective: they work after something has already happened. The real gap is during the event, in the live moment, where coordinators are juggling multiple conversations at once, and no existing tool helps prioritize or prompt them at that moment.
THE LEARNING: COMPETITOR ANALYSIS
Action items only get identified after the fact, which means it's often too late to act, leading to missed opportunities and delayed resolutions. This is exactly the gap our project aims to address: supporting coordinators in the real-time moment, not before or after.
THE PILOT
Before running formal interviews, we ran a pilot study, five interviews across two academic, one social, one corporate, and one wedding event coordinator, to learn how coordinators communicate as an event's flow evolves in real time.
Pilot Study
THE FIELD OBSERVATION
The Design of Everyday Things by Don Norman
I encouraged the team to do field observation because it would show us the gap between what coordinators say they do and what we actually see them do on the ground.
On-site during Field Observations
We ran two field observations, an Academic Research Showcase and a UX Conference. Updating every movement and interaction we saw the event coordinator make in our internal team Slack thread as it happened.
THE LEARNING: 5 PILOT INTERVIEWS & 2 FIELD OBSERVATIONS
Coordinators aren't sitting at a desk managing things. They're constantly moving between people, locations, tasks, and tools, all while trying to keep everyone aligned
That alignment isn't simple. They're dealing with multiple stakeholders at once, each with different needs, priorities, and expectations.
Coordinators are rarely playing just one role. In a single moment, the same person might be the planner, the host, the decision-maker, and the problem-solver.
That pushed us to update our research questions:
In unpredictable moments, how do they adapt their communication strategies to keep events flowing?
We also made a deliberate choice to focus on live, in-the-moment coordination. Live events are high-stakes: decisions are irreversible, failures are simultaneous and compounding, and decision-making happens under chaos. They're also high impact: reputation and client trust are on the line, and execution affects future revenue and bookings.
USER INTERVIEWS
I contributed to conducting the user interviews. We completed nine in total: one Executive Assistant, one Concert Manager, one Wedding Assistant, two Event Managers, two Event Coordinators, one Special Events Staff Lead, and one Attendee Services Coordinator.
Participant 08
Participant 09
Participant 03
Four consistent friction points emerged across those interviews, the subject of the next section.
03 — THE SYNTHESIS
This is the part of the research where my individual contribution mattered most, not in generating the insight alone, but in refusing to let the team settle on the first, easiest read of the data. I led the synthesis of the research findings using FigJam AI to quickly surface initial patterns, then went back through its output myself to double-check it. Everything I learned was a surface-level problem: tools failed, networks dropped, people did not answer, and messages were scattered across platforms.
Research synthesis I quickly led
I didn't think that was the real story. I believed every one of those was a symptom of something deeper, not the problem itself, and I told the team so. I used an analogy: a sick patient's symptoms describe how the sickness shows up, but the real problem is what's happening beneath those symptoms.
The team pushed back. Their position was that the pain points were the real problem. I didn't stay quiet. I tried the research ladder, asking “why” and “how,” to dig past the pain points toward what coordinators were actually trying to accomplish: getting their job done.
My 1st Artifact I Presented to the Team
Still pushback. So I did the analysis on my own and brought it back to the team a second time. A teammate pushed back again, this time asking me to back the argument with actual interview data, specific participants, specific quotes.
My 2nd Artifact I Presented to the Team
So I did it a third way: I pulled in every relevant participant quote and explained not just coordinators' workflow, but how their thinking actually works while coordinating across four constraints (cognitive, physical, social, and tool) each backed by direct quotes from interviews.
My 3rd Artifact I Presented to the Team
GETTING TO ROOT CAUSE, TOGETHER
The team still wasn't fully convinced, but they wanted to explore the thinking further. Together, we expanded the constraint framework to also include time, environment, and availability. That reorganization surfaced four findings, each backed directly by interview data.
We then built a mind map to trace how these four findings connected to each other, and every line led back to the same place: cognitive capacity. Multi-role management and the event environment consume that capacity; time pressure and unreliable communication tools limit coordinators' ability to keep information flowing, which then consumes still more capacity.
The Team's Mindmap to Find the Root Cause
That's how we agreed the real problem was:
This wasn't a theoretical loop. It's the same pattern across all four findings above: something in the coordinator's environment interrupts or consumes their cognitive capacity, which degrades the information reaching them, which consumes still more capacity. Any solution we designed would need to work within that limited capacity, not add to it.
The Vicious Cycle We Defined
Participants confirmed it themselves. 8 out of 9 told us they didn't want to change how they communicate; they wanted a better tool to do it with.
Participant 04
Cognitive overload as root cause opens up the design space: not asking coordinators to work differently, but giving them a tool that carries some of the cognitive load for them.
04 — THE DESIGN CHALLENGE
The research aligned with our design principle of AI role in the solution: not as a final decision-maker, but as a second brain, augmenting coordinators' cognitive capacity to handle unstable information so they can be even more effective in high-stakes, real-time situations.
WHAT THIS MEANS FOR THE DESIGN
WHERE THE PROJECT STANDS NOW
Research is complete. The team has moved into the design phase, translating this reframed problem and design challenge into concrete concepts and prototypes. This case study will be updated as the design phase progresses.
NEXT PROJECT