SilverJoy Planner HK

A plan suggestion for family with children
Friendly form input, avoid long paragraph with AI
Plan for vegetarian and seniors

Inspiration

We were inspired to build SilverJoy Planner HK after noticing that many elderly and families in Hong Kong struggle to plan accessible, enjoyable outings. Most trip planners lack essential local context—like barrier-free routes, real-time crowd data, elevator and restroom information—that are crucial for safety and comfort. Our team wanted to use AI to create an inclusive, stress-reducing trip planner for all.

What it does

SilverJoy Planner HK is an AI-powered web assistant that creates daily itineraries for families, seniors, and visitors with accessibility and comfort in mind. It integrates live data (weather, points of interest, accessibility options, restrooms, lifts, and cost) to generate step-free, low-stress, and weather-adaptive travel plans. Each personalized route includes stop suggestions for meals, rest, and emergencies, tailored for safe and enjoyable Hong Kong experiences.

How we built it

Streamlit for rapid prototyping of the web interface

OpenAI/LLM APIs for intelligent trip plan generation with contextual understanding

Public datasets from data.gov.hk, Google Maps, and HK Transport for POIs and accessibility info

Custom Python pipelines for data cleaning and processing

Cloud deployment for real-time use and easy access

Challenges we ran into Integrating messy, incomplete public datasets (missing or inconsistent accessibility and POI info)

Getting the LLM to accurately prioritize accessibility features like stair-avoidance, elevators, and restroom proximity

Balancing the need for up-to-date information (live weather/crowd data) with fast response times

Designing an easy-to-use interface suitable for seniors as well as younger tech users

Accomplishments that we're proud of Delivered a functioning, context-aware itinerary engine in less than a week

Successfully merged real Hong Kong accessibility data with smart AI reasoning

Received positive feedback from elderly and teacher test users for usefulness and user-friendliness

What we learned

We learned the importance of deep local context when applying AI to real-world problems, the technical challenges of combining public data with LLMs for practical solutions, and UI/UX principles for building tools usable by people of all ages.