WHY - The Problem
OneStop consolidates NTU's fragmented university systems into a single AI-powered interface. We identified a clear entrepreneurial opportunity: students waste 15 minutes daily navigating disconnected platforms.
WHY - Opportunity Recognition
Through customer development and user research, we identified four core pain points in the current university digital ecosystem.
Students juggle between STARS, NTULearn, Campus Map, and countless other platforms, each serving limited functionality.
Tools require navigating between multiple sites, creating friction during time-sensitive situations like finding exam venues.
Students must remember which tools to use, manage constant interface-switching, and handle relentless re-authentication.
Students must actively search for information, risking missed deadlines or last-minute schedule changes.
Shows bus locations and ETAs. No integration with class schedules - students must manually check if they'll make it to class.
Course registration and timetable. Doesn't show building locations or walking times between venues.
Course materials and announcements. Separate login, no calendar sync, notifications buried in email.
Building locations without real-time data. No walking time estimates or integration with schedules.
Open Omnibus > Check bus time > Open STARS > Find room > Open Maps > Get directions > Hope you make it
"Where's my next class?" > Get room, walking time, and next bus in one response
WHAT - Value Proposition
OneStop consolidates essential university services into three core modules, delivering clear value through innovation and simplification.
Our core value proposition: Replace 100 clicks with one conversation. The AI chatbot synthesizes data from 2,000+ locations and 5,000+ courses to deliver instant, contextual answers.
WHAT - Key Resources
The chatbot is our primary differentiator - transforming complex multi-platform queries into simple natural language conversations.
Ask questions in plain English. "Where is LT19?", "When's the next bus?", "Find halal food nearby" - the AI understands context and delivers instant answers.
Recognizes 60+ food keywords, 35+ cuisines, and 55+ food brands. Real-time results from Google Places API with opening hours and ratings.
Maintains conversation history and injects relevant NTU data to provide accurate, personalized responses based on your schedule and location.
Bus arrivals, class schedules, exam venues, food options - everything accessible through one conversation instead of five different apps.
HOW - Key Activities
Our key activities focus on creating value through intelligent features that prioritize student convenience.
Proactive information delivery tailored to your day.
All course information consolidated in one place.
Real-time campus data at your fingertips.
Everything hall-related in one place.
HOW - Technical Infrastructure
Modern technology stack designed for reliability, speed, and great user experience across web and mobile.
GPT-4o-mini powers conversational AI with context-aware responses and 500-token limit per query.
Singapore's official mapping service for location search, geocoding, and walking route calculations.
Real-time food discovery with ratings, opening hours, and nearby restaurant information.
WHO - Customer Segments & Key Partners
Following the Entrepreneurship Ecosystem Framework, we engage stakeholders across the entire value chain.
Primary consumers who save time and reduce stress navigating campus.
Government/Leader stakeholder - bridges smart campus initiative with reality.
Service providers who integrate operations and reach students effectively.
NTUSU and hall councils as distribution channel for announcements.
WHO + HOW - Business Model Canvas
Mapping our solution to the Business Model Canvas framework taught in the course.
Mobile app (Android via Capacitor), Progressive Web App, Campus marketing through NTUSU, Hall orientation programs
OpenAI API costs, Cloud hosting, Development resources, Marketing & distribution
Pricing: $0.15/1M input tokens, $0.60/1M output tokens
Estimate: 5,000 DAU x 3 queries x 500 tokens = 7.5M tokens/day
Monthly: ~225M tokens = $135-270 (with caching)
Pricing: $17/1,000 requests (Place Details)
Estimate: 2,000 food searches/day x 30 days = 60,000/month
With caching: ~3,000 unique requests = $51
Frontend: Vercel Pro = $20/month
Backend: Railway Hobby = $5-20/month
Database: Supabase Pro = $25/month (8GB, 50GB bandwidth)
Pricing: Free for educational institutions
Limit: 250 calls/min (sufficient)
Note: NTU qualifies as public educational institution
Access to 33,000 students enables precise targeting. Food vendors, campus services, and local businesses can reach students at decision moments.
Canteen operators pay for promoted listings and real-time menu updates. Students searching for food see sponsored options first.
NTU sponsors the platform as part of smart campus initiatives, reducing operational costs for student services.
HOW - Execution Milestone
Following the Design Thinking methodology with Solution Design, Prototype, Pilot Testing, and Scale phases.
IMPACT
Measuring success through tangible outcomes for students and the university.
Enhanced quality of education through reduced friction. Students focus on learning instead of navigating systems.
Increased productivity for 33,000 students. Time saved translates to better academic outcomes and reduced stress.
GROUP REFLECTION
As students from Computer Science, Data Science and AI, and Mechanical Engineering, we approached OneStop with distinct perspectives. Our Computer Science members focused on system architecture, API integrations, and frontend design. The Data Science and AI perspective shaped how we designed the chatbot's natural language understanding and query routing. Our Mechanical Engineering background brought a user-centric design mindset, thinking about physical campus navigation and how digital solutions integrate with the real-world environment students move through daily.
This interdisciplinary collaboration taught us that entrepreneurship is not just about building products. It's about understanding real problems from multiple angles and pitching solutions that address genuine needs. The ET5131 module challenged us to think beyond our technical comfort zones.
Our biggest takeaway: technology is only valuable when it solves a genuine pain point. Every feature we prototyped started with a question. Does this actually help students navigate campus life better?