Design & research / 2025–present
Giving care workers more time for people, with an AI-assisted workflow shaped around their needs.
Ongoing · research & prototype
The 30-second version
Context
Hong Kong’s ageing population puts increasing pressure on geriatric care. BeeCare explores how AI-assisted workflows could give frontline staff more time for direct care.
The project began with the pressure of the impending silver wave on Hong Kong’s geriatric-care sector. My focus was the work around care: how people record, find, and pass on the information they need.
As co-founder and Lead R&D, I led user research and service design and translated social-work requirements into product features and high-fidelity prototypes.
Problem
Research pointed to administrative overload, fragmented information, and digital tools that were difficult for frontline workers to navigate.
Documentation eats up time meant for direct care.
Information is spread across paper folders and disconnected digital tools.
Existing systems can be difficult for frontline workers to navigate.
Through user research, we identified a recurring concern: social workers were caught in administrative cycles they experienced as “meaningless.” Documentation took time intended for direct care.
Information was scattered across paper folders and disconnected digital tools. Existing systems could also be difficult for middle-aged frontline staff to navigate.
Research
I combined professional input with a modified Delphi process to identify pain points and establish priorities.
I developed and carried out a research framework to ground the solution in professional needs and academic practice. A modified Delphi method helped establish consensus around pain points and priorities. Interviews with veteran social workers at Amazing Grace Care Services added detail about the realities of frontline work.
The wider research and service-design process included contextual inquiry, focus groups, user-journey mapping, and high-fidelity prototyping. These methods helped translate social-work requirements into product decisions.
Design decisions
The design responds to visual, physical, and cognitive strain through large text, generous targets, speaking-first input, and simple navigation.
Large text and high contrast, chosen for frontline users.
Generous 72px targets in the prototype.
Voice input aims to reduce the burden of writing records.
Linear steps and flat menus reduce navigation complexity.
I translated qualitative findings into technical and UX choices. The prototype uses a 26pt minimum font size and high-contrast colours to accommodate declining vision. Touch targets of 72px aim to reduce physical input strain and errors.
To address input fatigue, the workflow shifts from writing-first to speaking-first. Linear navigation paths and flatter menu hierarchies aim to help people stay oriented.
Process
Desk research → data collection → analysis → core functions → experience design → rapid prototyping.
The process moved from desk research and empirical data collection to analysis, defining the main functions, and designing the user experience. AI-assisted rapid prototyping helped visualise the proposed functions.
The process material below preserves the research and interface work behind the shorter overview.
Solution
The concept brings AI-assisted data entry, a shared workspace, and reusable outputs into one care workflow.
Turn spoken observations into structured care documentation.
Bring the information staff need into one place.
Use recorded information across the required outputs.
AI-assisted data entry: caregivers speak and capture moments, with AI proposed to turn that input into structured care records.
A centralised workspace: information is brought together so staff can access it without switching between disconnected tools.
Smart output and automation: the concept is to input data once and reuse it across the outputs staff need. This remains a proposed workflow rather than a claim of a deployed, validated operating system.
Intended impact
Less administrative work for social workers, and care knowledge that can stay with an organisation as staff change.
For social workers, the ambition is to turn documentation that takes hours into work that takes minutes.
For elderly-care centres, the aim is to retain client histories so that an experienced worker’s knowledge can help train the next generation rather than leave with them. The project received PolyU Capstone Project Funding Scheme support and Beijing Delegation funding support.
Reflection
Those who take care of others also deserve to be cared for.
To me, meaningful and impactful design — that can improve people’s lives — is immensely rewarding and needs to be done. In this case, those who take care of others also deserve to be cared for.
Thank you for reading this far.
FaccuracyMaking credible information easier to find, understand, and question.Read the case study ↗