SNU Now

An AI-Powered Social Map for Campus Life

Project Proposal | Software Development Principles and Practice, 2026 Fall

01 Team Information

• Team name: Bonnie and Claude / Team 9
• Members: 김태현 · 안진영 · 윤유상 · 함재현

02 Project Name

SNU Now is an AI-powered social map for SNU's Gwanak campus. Students use the map and chatbot to find event companions, arrange meetups with friends, and access campus services.

03 Target Customers

• Students: SNU students want to meet friends between classes or attend events with new companions. Finding people with shared interests and compatible schedules takes effort.
• Event organizers: Companies promoting campus recruiting events need to reach interested SNU students and turn event discovery into participation.

04 Motivation

Students currently piece together schedules, locations, and event information across services. SNU Now connects discovery, coordination, and action through three core values.

• Offline Meetings: Friends can meet between classes without missing their next commitments.
• Event Participation: Students can find companions with shared interests for events and activities afterwards.
• Everyday Convenience: Students can complete campus tasks by describing what they want in chat.

05 Main Features & AI Utilization

• User inputs: Profiles contain names, departments, admission years, and editable interest hashtags. Students may enter a timetable, upload its image, or state free times. They describe desired companions in a prompt.

A. Offline meetings

• Friend meetups: Friends appear as avatars with statuses on the map. After a meetup request is accepted, the app suggests an activity, campus venue, and time that fit both schedules and travel times. The agreed plan becomes a shared quest that only participants can see, track, and open for directions.

B. Event participation

• Official events: The app imports selected departmental notices and organizer submissions, then filters events by relevance and importance. Map cards show times, venues, eligibility, and sources.
• Private events: Students can create events manually or upload poster images to extract event details. They can edit or delete these events, which appear only on their own map.
• AI party matching: A party is a group attending an activity together. AI uses interest hashtags, companion-seeking prompts, and availability to recommend open parties or form groups among students seeking companions. Suggestions explain shared interests and overlapping free time. Students can invite friends. They can edit interest hashtags and give feedback on suggested companions to refine future matches.
• Organizer tools: Map listings, QR badges, in-app rewards, and participation counts could support paid promotion and participation management. Each account can claim one reward per event.

C. Everyday convenience and AI chatbot

• Campus services: Students can check dining menus, study-space information, and library seat availability, and see shuttle stops and vehicles on the map. Chat recommends places that fit travel time and upcoming classes. Live data and library reservations depend on provider access.
• AI chatbot: Students can use all implemented app functions through chat. For "Plan a café visit with them," the chatbot retains the selected event and party, clarifies missing details, and calls app tools in sequence. Code checks permissions, schedules, travel times, party membership, and reward eligibility before execution.

D. Privacy and location-sharing controls

• Friends: The acceptance screen explains that accepting starts mutual location sharing. Users can pause or limit sharing.
• Parties: Members share locations until leaving, which immediately revokes party access. Friend permissions remain separate.
• Protection: The app hides off-campus locations and user-defined private zones such as dormitories. Servers enforce blocking and permissions, limit retention, and restrict AI to authorized information.

06 Initial Scope & Extensions

• Initial release: Android users will sign in through SNU SSO, if permitted, or SNU-managed Google accounts.
• Data: We will check access to departmental notices, SNU menus, and Kakao Maps early. Demos will use clearly labelled sample data when live sources are unavailable.
• Priority stretch goal: After core chat works, we will prototype chat autocomplete like GitHub Copilot's code completion. Suggestions will use event, party, and conversation context to reduce typing.

07 Devices Needed

We request one Samsung Galaxy S22 and one Galaxy S23 for invitation, location sharing, and QR tests.

08 Testing & Demo

• Main demo — Event Participation: On a recruiting-event card, a student asks chat, "Find people interested in AI careers to attend this event with me." They join a recommended or newly formed party, locate members through temporary location sharing, and attend together. After the event, they continue the conversation at a café recommended by chat.
• Demo — Offline Meetings: A student asks chat to arrange lunch with a friend at a campus dining hall. After acceptance, both receive a shared meal quest fitting their schedules.
• Demo — Everyday Convenience: A student checks dining menus and study spaces, views shuttle stops and vehicle locations on the map, and asks chat to check library seats and make a reservation.
• User evaluation: From the first iteration, SNU students will compare the three demos with manual workflows using the same information. We will measure completion rates, task times, and satisfaction, and record chatbot response times.
• AI evaluation: Using the same candidate pool, we will assess whether matches satisfy requests for similar interests or diverse backgrounds. We will compare the chatbot's tool selection accuracy and completion of multi-step requests with and without conversation context. We will also measure image extraction accuracy and correction effort.
• Reliability: We will test sharing revocation, unauthorized access, duplicate QR claims, map speed, and battery use.