Inspiration
NeighbourAssist was inspired by one team member's firsthand experience volunteering at Meet-the-People Sessions (MPS) in Singapore.
During these sessions, he encountered residents, particularly elderly individuals, seeking assistance with financial difficulties, housing arrangements, CPF matters, medical expenses and other everyday challenges.
A recurring problem was that some residents struggled to understand which government assistance schemes were available or which agencies they should approach. Others found digital government services difficult to navigate.
He also witnessed queues lasting over an hour, with elderly residents waiting late into the evening. On one occasion, he even escorted a frail elderly resident home after a session.
These experiences made us question something:
If government assistance already exists, why can finding the right help still be so difficult?
We realised that while Singapore has numerous assistance schemes and support channels, navigating them can be overwhelming, especially for elderly residents unfamiliar with digital services.
When our teammate shared his experiences and proposed the idea, we decided to develop NeighbourAssist for Neighborhood Hacks 2026.
Our goal was not to replace government agencies or community volunteers, but to make the first step towards receiving assistance simpler, clearer and more accessible.
What It Does
NeighbourAssist is a free, bilingual, accessibility-focused website that guides Singapore residents towards relevant government assistance channels through simple, interactive questionnaires.
Instead of requiring users to understand government policies or identify the correct agency themselves, NeighbourAssist asks straightforward questions about their circumstances and provides relevant guidance.
Nine Assistance Categories
Our website currently covers nine categories:
- Everyday Expenses — Finding financial assistance for essential living costs, including urgent ComCare enquiries.
- Housing & Rental — Guidance on HDB rental matters, downsizing and housing concerns.
- CPF & Retirement — Understanding CPF-related enquiries and retirement support.
- Medical Costs — Finding information about healthcare subsidies and financial assistance.
- Neighbourhood Issues — Directing residents towards appropriate municipal feedback channels.
- Fines & Appeals — Identifying responsible agencies and understanding official appeal or payment procedures.
- Work & Employment — Guidance on salary disputes, dismissal, workplace fairness and employment-related concerns.
- Family & Social Support — Finding family counselling, caregiving and other community support services.
- Not Sure — Helping users who have received confusing government letters, need to follow up on applications or simply don't know where to begin.
After completing a questionnaire, residents receive a recommended official starting point, practical next steps, useful information to prepare and available contact methods.
Accessibility Features
Accessibility was a major consideration during development. We incorporated:
- English and Simplified Chinese support.
- Adjustable text sizes for improved readability.
- Caregiver mode, which adapts questions for someone assisting another person.
- Mobile-friendly layouts and simplified navigation.
- Optional MPS preparation, including printable notes and guidance on finding an MP.
Privacy and Safety
We designed NeighbourAssist to help residents navigate assistance channels without collecting sensitive personal information.
The website does not require accounts or collect NRIC numbers, medical records, banking details or personal documents.
NeighbourAssist is an independent prototype, not an official government service. It does not submit government applications, determine eligibility or guarantee assistance.
How We Built It
We developed NeighbourAssist using the following technologies:
- React — Building the interactive user interface.
- TypeScript — Implementing the questionnaire and recommendation logic.
- Vite — Frontend development and production builds.
- GitHub — Version control and source-code management.
- Deployxa — Hosting our live website.
- Node.js Test Runner — Testing our recommendation logic and important application behaviours.
1. Designing the Recommendation Engine
At the centre of NeighbourAssist is a rule-based recommendation engine.
Instead of relying on AI to make potentially unreliable decisions about government assistance eligibility, we designed conditional questionnaire pathways that direct users towards relevant official sources based on their selected answers.
For example, someone who urgently cannot afford basic necessities receives a shortened questionnaire and guidance on contacting ComCare, rather than being required to answer unnecessary questions.
Similarly, our housing questionnaire distinguishes between different living arrangements to avoid providing the same recommendation to every resident.
2. Researching Official Assistance Channels
We researched Singapore government assistance channels and incorporated official agency links, contact information and explanatory guidance into our results.
Our aim was to provide residents with clear instructions while ensuring that official agencies remained responsible for determining eligibility and processing applications.
3. Building an Accessible Interface
We implemented bilingual translations, caregiver-specific wording, responsive layouts, adjustable text sizes and an optional MPS preparation feature.
The preparation feature generates printable notes locally from the user's selected answers without transmitting their case information to a server.
4. Testing and Deployment
We developed automated tests covering major recommendation pathways, translations, urgent situations and accessibility-related behaviours.
We also used GitHub to manage development and Deployxa to publish NeighbourAssist as a working website.
5. AI-Assisted Development
We used AI tools, particularly ChatGPT, extensively to support technical implementation. Our approach, responsibilities and lessons learned are explained under Challenges We Faced.
Challenges We Faced
1. Simplifying Complex Government Information
Government programmes often have different eligibility requirements, application procedures and responsible agencies.
Our biggest challenge was transforming this information into guidance that ordinary residents could understand without oversimplifying important details.
We addressed this by using plain-language questions, distinguishing between different circumstances and directing users towards official sources for final decisions.
2. Designing for Elderly Residents
Our target audience includes people who may be unfamiliar with technology.
We had to consider readability, confusing terminology, questionnaire length and mobile accessibility.
These considerations required us to evaluate whether the interface would be understandable and practical for residents with different levels of digital literacy.
3. Avoiding Misleading Recommendations
We wanted NeighbourAssist to be helpful without suggesting that it could determine eligibility, overturn government decisions or guarantee assistance.
This was especially important when handling urgent financial needs, legal deadlines and potentially unsafe situations.
We incorporated relevant warnings and prioritised official agency and emergency channels where appropriate.
4. Supporting Caregivers
We realised that not every user would be seeking assistance for themselves.
Family members and caregivers might use NeighbourAssist on behalf of elderly relatives.
This led us to introduce a dedicated caregiver mode and adapt the questionnaire wording to reflect the person receiving assistance.
5. AI-Assisted Development
Artificial intelligence played a significant role in developing NeighbourAssist.
As a team with limited experience in web development, we relied heavily on AI tools, particularly ChatGPT, throughout the project.
AI assisted us with planning the application's architecture, generating and refining code, implementing questionnaire logic, designing the interface, debugging errors, writing automated tests, and researching government assistance channels.
Our role involved identifying the real-world problem, defining the project's objectives, deciding which features to prioritise, reviewing AI-generated work, testing the website and refining the user experience.
Although much of the technical implementation was AI-assisted, we remained responsible for the project's direction, decisions and final output.
This experience taught us how to use AI as a practical development tool while recognising the importance of independently reviewing its suggestions, particularly when dealing with information that could affect vulnerable residents.
Without AI assistance, building a functional prototype of this scope within our limited timeframe and experience would have been significantly more difficult.
What We Learned
Building NeighbourAssist taught us that a meaningful solution does not always need to be technologically complicated.
Initially, the problem appeared to be about access to government assistance. However, as we developed the project, we recognised that understanding where to begin can be a major challenge in itself.
We also learned the importance of designing for people whose needs and technological familiarity may differ from our own.
A feature that seems intuitive to us might be confusing to an elderly resident. This reinforced the value of clear language, accessibility and human-centred design.
On the technical side, we gained experience in:
- Building interactive applications using React and TypeScript.
- Designing conditional questionnaires and rule-based recommendation systems.
- Managing bilingual translations and responsive interfaces.
- Writing automated tests to check application behaviour.
- Deploying and maintaining a publicly accessible website.
- Balancing functionality, accuracy, accessibility and privacy.
Most importantly, we learned how to turn a problem observed firsthand in our community into a functioning prototype.
What's Next?
NeighbourAssist is still a prototype, and there are several areas we would like to improve.
Our next priority would be conducting usability tests with elderly residents, caregivers and community volunteers using fictional scenarios.
We would use their feedback to simplify confusing questions, improve accessibility and identify missing assistance pathways.
Our future plans include:
- Expanding language support to better serve Singapore's multilingual population.
- Improving accessibility, particularly compatibility with screen readers and assistive technologies.
- Maintaining accurate guidance by regularly reviewing official government resources.
- Exploring community partnerships with social-service professionals and volunteer organisations.
Ultimately, we hope NeighbourAssist can complement the work of government agencies, social workers, community volunteers and MPs.
Our vision is simple: nobody should struggle to find help simply because they don't know where to begin.
Finding the right help shouldn't be hard.
Try NeighbourAssist
- Live Website: neighbourassist.deployxa.online
- GitHub Repository: NeighbourAssist on GitHub
Built With
- deployxa
- github
- react
- typescript
- vite
Log in or sign up for Devpost to join the conversation.