A student-led platform for sharing failure stories and building resilience among university students.
🏆 NUS-funded Project — S$5,000
🌏 Five-country International Team
🎓 NUS Asian Undergraduate Symposium (AUS) 2026
Project Lead & Presenter: Shuyue Quan / 全书阅
Open the live browser experience
Developed through a five-country international team during NUS AUS 2026, with English used as the primary language for project discussion, coordination, and presentation.
University students are constantly exposed to stories of success, while failure is often hidden. Fail Forward Gallery creates a safer space for students to share setbacks, learn from others, and access practical support.
Current status: active six-month implementation following the AUS-funded concept stage. The public main branch contains the working bilingual browser experience; the account, ownership, and human-moderation foundation is implemented and tested locally as V0.2 work in progress.
Fail Forward Gallery is an anonymous storytelling and peer-support platform designed first for university students in China. It gives students a place to talk about academic setbacks, job-search rejection, financial pressure, loneliness, and difficulty fitting in without attaching those experiences to a public identity.
The project combines a searchable story archive with choice-based responses, practical resources, and author follow-ups. It is a peer-support and educational project, not a clinical service, and it does not promise that every story must end in success or recovery.
Achievement is highly visible on campus and online; disappointment is usually private. Students can therefore experience a common setback as evidence that they alone are falling behind.
Fail Forward Gallery was created to challenge that imbalance. Instead of treating failure as a personal defect, the platform makes room for honest, unfinished stories and connects emotional validation with context-specific support. Its China-first design reflects the pressures surrounding examinations, postgraduate admissions, recruitment, family expectations, the urban–rural divide, and studying away from home.
- Selected for the NUS Asian Undergraduate Symposium (AUS) 2026.
- Awarded S$5,000 in project funding by the National University of Singapore.
- Developed by an international undergraduate team spanning five countries.
- Supported by a six-month implementation period to move from concept to a safer, testable product.
- Designed as an open and adaptable model that can later support collaboration across universities in Asia.
| Product area | What it enables | Current state |
|---|---|---|
| Failure Story Archive | Browse anonymous stories by thematic board, tag, response preference, and search | Working |
| Choice-based Comments | Authors choose encouragement, similar experiences, gentle advice, or no comments | Working prototype |
| Advice Hub | Open card-based, context-specific guidance linked to common student setbacks | Working |
| Follow-up System | Original anonymous authors can add an update without being pressured to present a positive outcome | Working prototype |
| Bilingual Experience | Chinese-first interface with an English option | Working |
| Account & Story Ownership | Private mobile-based account foundation separates ownership from public anonymity | V0.2 work in progress; not yet on main |
| Human Review Workflow | New stories remain private until an authorised reviewer approves them; decisions include reasons and audit events | V0.2 work in progress; not yet on main |
| AI-assisted Safety Screening | Risk classification before human review | Planned; no production AI service connected |
| Ranking & Discovery | View-, response-, and editor-informed discovery with anti-abuse safeguards | Planned |
| Pop-up Exhibition | Physical story gallery and facilitated reflection activities | Planned pilot component |
The public browser experience and local V0.2 service are real working software, but the project is not yet a production mental-health platform. Real SMS delivery, production cloud storage, AI safety screening, review-notification email, regional crisis-resource operations, and shared public deployment still require implementation and operational review.
Project Lead & Presenter
- Led the development of the problem framing, product concept, and implementation roadmap.
- Coordinated an international team across five countries, using English for project discussion, collaboration, and presentation.
- Conducted and synthesised student interviews to connect lived experiences with product decisions.
- Managed product iteration from the symposium concept through the funded implementation stage.
- Presented the project in English at NUS AUS 2026.
- Led post-AUS implementation, including the transition from a static prototype toward accounts, content ownership, and accountable human moderation.
- Maintained the product’s focus on anonymity, user control, psychological safety, and culturally grounded design.
We conducted interviews with university students to understand how they experience academic failure, social isolation, recruitment pressure, and the expectation to present success online. We do not publish raw interviews or identifying details in this repository.
The research is used as a product input rather than decoration:
| Research insight | Product decision |
|---|---|
| Students often hide setbacks because they expect judgment or do not want a difficult moment attached to their identity. | Public posts and responses are anonymous, while private account ownership enables follow-ups, withdrawal, and review status. |
| Not everyone wants advice when sharing something vulnerable. | Authors choose their preferred response mode, including a no-comments option. |
| Generic encouragement can feel disconnected from academic, financial, family, or cultural realities. | Stories, tags, and Advice Hub content are organised around specific contexts and lived categories. |
| A setback does not always resolve neatly or positively. | Follow-ups allow “still difficult,” “no major change,” small steps, improvement, or a preference not to disclose. |
| Students wanted easier ways to find relevant experiences without flattening them into a single narrative. | Board filters, shared tag data, search, and planned responsible ranking improve discovery while preserving context. |
These findings continue to shape tag design, discovery mechanisms, interface refinement, moderation rules, and the kinds of stories and support the platform makes room for.
The current product includes a Chinese-first home experience, a filterable anonymous archive, story details and follow-ups, a card-based Advice Hub, a guided submission flow, private “My Stories” ownership, and an authorised human-review queue.
Readers never see a phone number, email address, internal account identifier, or cross-story public profile. Private account ownership exists only so an author can recognise and manage their own contributions.
New server-backed submissions stay private until a human reviewer approves them. The product separates automated risk signals from the authority to publish: AI may eventually reject clear violations or route ambiguity to review, but only a human reviewer can approve public content.
The interface avoids recovery scores, success rankings, streaks, and forced positive conclusions. “No change” is treated as a valid follow-up.
Authors choose tags, response boundaries, follow-up visibility, and—within the planned production workflow—withdrawal and appeal options.
The emotional experience of failure can cross borders, but its causes and consequences differ. Boards, tags, advice, and moderation preserve the context of each story rather than reducing it to generic motivation.
The project is in active V0.2 implementation. The public main branch currently contains the accepted V0.1 browser experience. The V0.2 account and moderation layer is implemented and tested in the current local development branch and will be published after acceptance and a history-safe release.
- Chinese-first interface with English switching.
- Story Archive with board, tag, response-mode, and text filters.
- Shared tag taxonomy across filtering and story submission.
- Anonymous story detail, comments, no-comment boundaries, and a Day 30 follow-up example.
- Guided 150–200 unit story submission with English-word and CJK-character counting.
- Advice Hub, accessibility foundations, mobile layouts, and demo reset controls.
- Mainland China mobile-number verification flow with development-only terminal codes.
- Private HttpOnly sessions and pseudonymised phone storage.
- Server-backed story submission and private “My Stories” status.
- Role-protected human-review queue, approve/reject reasons, and immutable audit events.
- Public API access only to human-approved anonymous stories.
- Production SMS provider and regional PostgreSQL deployment.
- Server-backed comments and follow-ups using the same ownership and review model.
- Compliant AI-assisted risk classification with human-only publication approval.
- Notifications, appeals, withdrawal, reporting, and moderator wellbeing procedures.
- Responsible ranking and discovery with rate limiting and anti-manipulation safeguards.
- Pilot evaluation, pop-up exhibition materials, and a reusable implementation toolkit.
The accepted product and architecture foundation is documented in V0.1. V0.2 documentation will be published with the corresponding service code.
The stack is intentionally small so the team can inspect and improve the full product:
- Frontend: semantic HTML, CSS, and vanilla JavaScript
- Public main branch: static HTML, CSS, and vanilla JavaScript
- V0.2 development: Node.js and SQLite
- Product structure: bilingual interface and a shared story/tag model; modular authentication and moderation adapters are in V0.2 development
- Testing: Node.js syntax checks and built-in test runner
No React, Vue, frontend build tool, or third-party UI framework is required.
Run the current build locally
python3 -m http.server 4173Open http://127.0.0.1:4173.
This runs the accepted static browser experience on the public main branch. V0.2 Node.js service instructions will be published together with the accepted V0.2 source and tests. Development verification codes are terminal-only simulations; they are not real SMS messages.
- Do not put real names, student IDs, contact details, interview transcripts, or identifiable research material into sample stories.
.env, local secrets, SQLite files, and logs are excluded from Git.- This project provides peer support and education; it is not a substitute for clinical care or emergency services.
- Production launch requires verified region-appropriate crisis resources and a documented escalation process.
Shuyue Quan / 全书阅
Project Lead & Presenter, Fail Forward Gallery
NUS Asian Undergraduate Symposium 2026
Failure is not a fixed ending. Stories can continue, change, or remain unresolved.