myServices

Inspiration

Growing up and being a part of a migrant (citizens and non-citizens) community, we saw firsthand how moving to a new country or city creates a gap between available support and the people who actually use it. Many migrants and new residents in Melbourne don’t struggle because services are missing, but because they’re hard to understand, scattered across different sites, and written in complex language. We wanted to reduce that friction and make support easier to access from day one of their journey.

What it does

myServices is an evidence backed multilingual mobile navigator bridging the gap between immigrants and the complex Australian support ecosystem. Specifically built for newcomers and migrants in Victoria, it removes the knowledge barrier and allows ease of access.

Instead of forcing users to search for services they might not know, myServices uses a questionnaire onboarding flow. By capturing basic details, visa type, age, income, and location, the app generates a tailored list of services categorised by eligibility:

Likely Matches: Services that closely align with the user's profile.
Worth Checking: Potential supports that require further verification.

Evidence: myServices directly addresses what researchers in Melbourne identify as the "legal, financial, and language barriers" that cause migrants to struggle with "difficulty in navigating the health care system" (Biddolph et al., 2025). It moves away from the usual self-search system that often leads to an "ineligibility to access Medicare, resulting in differential health care" for vulnerable migrants.

AI Situation Search

Life doesn't always fit into a search bar. Our AI feature allows users to describe their situation in their own words (e.g., "I lost my job and I'm worried about rent"). The app infers intent and maps the person to specific service categories. Results are presented as cards that provide:

Extra service details.
A guide for further navigation.
Direct links to official Victorian services.

Evidence: myServices adopts a "human-centred design" approach to "uncover unmet community needs" that traditional systems often miss (State of Victoria, 2022). This eliminates the "knowledge gap" that the University of Melbourne identifies as a primary driver of social exclusion, ensuring personal crises are immediately translated into actionable service roadmaps (University of Melbourne, 2022).

Multilingual Support

Built for a diverse Victoria, the app currently supports 12 major languages and Right-to-Left (RTL) layouts from the first screen. Supported languages include Arabic, Mandarin, Sinhala, Somali, Hindi, and more.

Evidence: By supporting 12 major languages, myServices caters to the 27.6% of Victorians who speak a language other than English at home (ABS, 2021). Furthermore, alignment with the Multicultural Victoria Act 2011 ensures that "all individuals in Victoria are equally entitled to access opportunities," removing linguistic barriers.

Plain-Language Clarity

We stripped away “government speak,” replacing bureaucratic jargon with simple, trustworthy explanations of eligibility and requirements.

Evidence: By replacing bureaucratic "government speak" with clear terms, myServices addresses the "low health literacy" and "challenges navigating unfamiliar systems" affecting the 77.7% of overseas-born Victorians from non-English speaking countries (Department of Health, 2023). This fulfills the Victorian Government’s mandate for "clear, direct and easy to understand" communication (DFFH, 2022).

Offline-Aware Resilience

Recognizing that those in crisis often have unreliable data access, the app features:

Smart Caching: The service catalog and previous AI search results remain accessible without a connection.
Graceful Degradation: If the server is unreachable, the app utilizes local storage to provide eligibility estimates, clearly flagging when data is served from the cache.

Evidence: By prioritizing "offline-aware" caching, myServices addresses the "digital divide" where low-income Victorians often face a "mismatch of digital services" (University of Melbourne, 2022). It ensures those in crisis maintain "continuous access to critical information," supporting the Victorian Government's commitment to "digital inclusion" (State of Victoria, 2021).

How we built it

Frontend

React Native (Expo) with TypeScript
React Navigation for the onboarding → questionnaire → results flow
NativeWind for consistent, fast UI styling
i18next for multilingual support across key community languages (Arabic, Hindi, Korean, Tagalog, Vietnamese, Chinese, and others)

Backend

FastAPI serving a curated JSON catalog of services
REST endpoints for service listing and eligibility evaluation
Rule-based scoring system that separates:
- Likely eligible
- Worth checking
- This keeps results transparent and avoids overstating eligibility.

AI Layer

Google Gemini used for natural-language classification
Converts user-described situations into service categories
Fallback keyword matching ensures reliability if AI or network fails

Resilience

Local caching of service data
Offline eligibility evaluation using shared logic
Network detection via NetInfo to maintain consistent behaviour

Deployment

OpenAPI documented backend
Configurable API base URL for flexible deployment
Runs locally or on hosted infrastructure

Challenges we ran into

Real-world eligibility complexity: Services overlap, rules change, and eligibility is rarely binary. We addressed this by introducing “Likely” vs “Worth checking” instead of making hard claims.
Diverse language support: Our community spans many languages, but high-quality translations and reliable service information are limited. We prioritised key languages while keeping the system extensible.
Balancing AI with trust: We needed AI to help interpret user intent, without letting it make decisions that require accuracy and accountability.
Navigation consistency: Ensuring users always return to a stable home/results screen required careful handling of navigation state.

Accomplishments that we're proud of

Delivered a complete user journey: language selection → questionnaire → prioritised results → service details.
Built a multilingual-first experience tailored to real community needs.
Designed an honest UX that communicates uncertainty instead of overstating accuracy.
Implemented offline-aware functionality so the app remains usable in low connectivity.
Integrated AI in a controlled way that improves usability without replacing logic.

What we learned

Access is not just about providing information, but making it understandable and actionable.
Users need guidance on what to do next, not just a list of options.
Hybrid systems work best: Rules for reliability and AI for interpretation.
A focused, polished flow is more effective than a wide feature set.
Our assumptions changed quickly after engaging with real community perspectives.