A companion for cancer patients, designed for their lowest-energy moments.
MAYO CLINIC OBSERVERSHIP AWARDEE
Developed in collaboration with Mayo Clinic Oncology, ChemoBuddy is a patient-centered digital companion for people undergoing chemotherapy. It reduces cognitive load, offers just-in-time guidance, and helps patients manage symptoms and treatment between clinical visits.
Product
Healthcare app
My role
UX / Product designer
Timeline
10 weeks
Collaboration
Mayo Clinic Oncology
What Changed
Problem: The weight of a bad day
Chemotherapy patients are handed hundreds of pages of generic information at exactly the moment they have the least energy, focus, and emotional bandwidth to use it.
How might we help chemotherapy patients understand and manage treatment when their energy, focus, and emotional capacity are at their lowest?
What I Owned
Working within a seven-person team, these were the three pieces I owned end to end, from concept through clinical validation.
Research without a waiting room
I couldn't talk to chemotherapy patients, so I learned from three sources that covered each other's gaps:
Literature review
Before sketching a single screen, I wanted to know what the research already said about how chemotherapy patients process information and engage with digital health tools.
What are mHealth cancer apps doing well, and where are they falling short for patients managing treatment day to day?
How does symptom tracking affect patient-reported outcomes during chemotherapy?
Does digital patient education actually improve treatment adherence, or just information recall?
Where can a chatbot contribute safely and usefully in an oncology context, and where should it stop?
Reviewed 11 peer-reviewed oncology studies and digital health journals focusing on: mHealth cancer apps, symptom management and patient-reported outcomes (PROs), digital education tools and treatment adherence, and early experiments with chatbots in oncology.
Key findings
INSIGHT 1
Health literacy alone doesn’t guarantee patients follow a plan when information is complex or emotionally charged.
INSIGHT 2
Multimodal education (audio, video and simplified text) leads to significantly higher engagement than text-only.
INSIGHT 3
Chatbots work best as educational and supportive tools, not diagnostic engines.
Three findings that set the direction
How information is delivered matters as much as what it says. Multimodal education drove higher engagement than text alone.
Patients want one clear next step, and fatigue sets in fast. Existing tools were text-heavy, scattered across apps, and reactive.
Chatbots help most as educators, not diagnosticians.
Role-play can't replace real patients, so I made Dr. Umar's clinical review a required checkpoint for anything the app says about symptoms.
How I turned insight into decisions
I turned the findings into four "how might I" questions, then brainstormed features against them.
Low-fidelity wireframes
Before any visual design, I wireframed the five core flows (onboarding, dashboard, chatbot, symptom tracking and patient profile). I judged each screen against one question: could Sarah get through it on Day 4?
Onboarding: a calm start
My first concept opened on a dashboard. It was complete on paper, but picture Sarah on Day 4 facing a full screen of options before she can do anything. So onboarding became a conversation. A code from her care team pre-fills her details (or she sets up manually), then the app asks one question at a time and builds her plan. The dashboard comes after, not first.
Dashboard: a personalized treatment overview
It answers three questions at a glance: where am I in treatment, what's next, and what do I take today? A progress ring tracks cycles, and short messages mark the milestones, from "Every big journey starts with one small step" to "You're halfway there."
Education library: one lesson, three ways in
Every lesson exists as video, audio, and text, so patients choose what they have the energy for. The trade-off: each lesson is produced three times and kept consistent across formats.
Chatbot: support that knows its limits
I scoped the chatbot to education and support, and every chat opens with a reminder that urgent concerns go to the care team. It runs guided lessons in a format a patient picks and checks understanding with a short quiz. When a message signals danger, like a 104°F fever, it stops teaching and escalates: call the oncology team or go to the nearest ER, with directions.
Symptom tracking: making self-monitoring easier
Logging starts where patients already are, in the chat: "I have a severe headache." The app then opens a body map, narrows to the exact spot, asks for a 1-to-5 pain rating and saves the details. Severity indicators help patients tell "watch this" from "call someone." Dr. Umar reviewed the wording so it never reads as a diagnosis.
Patient profile: essential info in one place
Diagnosis, treatment, and care team (hospital and oncologist) sit on one screen, so contacts are always in the same place.
Impact
Designing for the hardest day paid off: in prototype testing, participants completed core tasks on their own, felt less anxious and described their symptoms more accurately, and the design was clinically validated by Dr. Umar.
©Gauri Rasote 2026. All rights reserved. Designed with love, one pixel at a time ;)