Microsoft Imagine Cup 2026

Hi, Lily

A low-cost AI voice companion robot designed for elderly people living alone.

Student Startup Team, Massey University, New Zealand

Discover How It Works Get In Touch

Loading 3D model...

Loneliness is a public health crisis

And it's growing. Millions of older adults live alone with limited daily interaction.

16.6%

Population of New Zealand aged 65+ (2023)

25.3%

Of those 65+, the percentage who live alone

59%

Of older adults feel lonely or socially isolated

Today's technology wasn't built for them

Smart Speakers

  • No long-term memory
  • Not designed for elderly speech
  • No care logic

Premium Companion Robots

  • Meaningful but too expensive
  • Price range: $175 - $9,000+

Apps & Portals

  • Screen-first design
  • High learning cost
  • Low emotional engagement

The Unfilled Gap: There is no scalable solution that is voice-first, memory-driven, affordable, and caregiver-friendly.

Memory creates companionship.
Affordability creates access.

Continuous Companionship

Remembers people, routines, and preferences over time. Lily becomes a familiar presence, not just a tool.

Emotion-First Dialogue

Designed to be warm, patient, and elder-friendly. Natural conversations without complex commands.

Quiet Wellbeing Signals

Inferred from natural conversation (non-medical). Detects pattern shifts in sentiment, routine, and engagement over time.

Ultra-Low-Cost Model

~$20 device + affordable subscription ($3-7/month). Hardware-as-a-Service makes companionship accessible to all.

"A familiar voice that remembers—and gently alerts family when patterns change."

Built on a modern, scalable Microsoft foundation

Azure OpenAI
Azure AI Speech
Azure AI Search
ESP32
Edge Device

See HiLily in Action

Watch our product demonstration and team pitch to understand how HiLily works and our vision for the future.

Product Demo

2 min

Live demonstration in a real home environment — voice interaction, long-term memory, and caring reminders powered by Microsoft Azure AI.

Team Pitch

3 min

Meet our team and discover the story behind HiLily — our mission to ensure no older voice is left unheard.

Real Voice Interaction

Long-term Memory Demo

Caring Reminders

Meet Our Team

A mission-driven team of builders and researchers from Massey University, New Zealand

Sarah Liu
Initiator

Sarah Liu

AI, Product & Integration

Information Science student at Massey University and the original initiator of the HiLily project. Involved throughout the entire workflow—from user requirement discovery, outreach and partner communication, to product design and hands-on engineering. Focuses on Microsoft AI stack integration, long-term memory pipeline, and end-to-end MVP integration.

Alex Luo
Core Dev

Alex Luo

Hardware, Prompt Engineering, Demo & User Research

Master's student in Information Science at Massey University with strong C/C++ and Python foundation. Owns the hardware workstream (ESP32-based voice device), contributes to prompt engineering, supports user needs research through requirement discovery and collection, and leads product demos to showcase the end-to-end MVP.

Jayden Chan
Core Dev

Jayden Chan

API Development & Pitch

Master of Information Science student at Massey University with a strong interest in AI and social impact. Develops product interfaces and APIs, and leads the pitch deck and project presentation to ensure the story and value proposition land clearly with judges and partners.

Faculty Advisor

Dr. Tong Liu

Dr. Tong Liu

Lecturer, School of Mathematical and Computational Sciences
Massey University

Specializes in Artificial Intelligence, Machine Learning, and Generative AI. Provides academic and technical guidance on AI methodology, model evaluation, and research rigor for the project.

Our design process started with our users

15

older adults interviewed
(ages 68-85)

12

family caregivers
surveyed

3

senior care organizations
consulted

Help us ensure no older voice is left unheard

— starting with real pilots.

Our Ask (Next Phase)

  • Guidance on responsible AI and privacy as we prepare for real-world eldercare pilots
  • Introductions to care organizations or NGOs interested in piloting a low-cost voice companion
  • Feedback on scalability and product-market fit as we move from MVP to early deployment

Entering a pilot-ready phase with a functional MVP and multiple iterations completed.

[email protected]

Massey University · New Zealand