National AI Student Challenge
Source: https://naisc.aisingapore.org
Archived: 2026-04-23 17:10
National AI Student Challenge
Skip to content
REGISTRATION PERIOD
5 JAN
16 FEB
REGISTRATION
CLOSES IN
Days
Hours
Minutes
Seconds
About National AI
Student Challenge (NAISC)
Empowering students by providing them with the platform and resources needed to accelerate their AI learning journey, fostering innovation and cultivating the next generation of AI users, practitioners, and creators equipped to tackle real-world challenges.
Why Join?
Build Real-world AI Skills
Gain hands-on experience tackling authentic industry and societal challenges.
Win Attractive Prizes and Internship Opportunities
Top participants can earn attractive prizes and internships offered by track owners.
Boost Portfolio and Career Prospect
Strengthen applications for polytechnic, university, internships, and future jobs.
Learn from Industry Experts
Receive mentorship, attend workshops, and access guided resources from AI professionals.
The National AI Student Challenge 2026 offers 8 Challenge Tracks, each with its own eligibility criteria and requirements.
Challenge Tracks
STEP
01
Please carefully review the deliverables and eligibility criteria for each of the Challenge Tracks below.
STEP
02
Local Tracks are exclusively available to students studying in Singapore, while the AI Ready ASEAN Track is open to regional students studying
in Singapore, Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines, Thailand or Vietnam.
STEP
03
Each Challenge Track has limited space. Registration is on a first-come, first-served basis and will close once all slots are filled. Secure your spot by applying promptly. Successful applicants will be notified by
23 February 2026.
8 CHALLENGES
Discover more about the eligibility and criteria for each challenge!
LOCAL TRACK 01
Certis
Learn More
Develop a multimodal AI advisor that analyzes security inputs and recommends context-appropriate responses based on standard procedures.
LOCAL TRACK 02
Huawei
Learn More
Create a smart multi-agent AI system that detects anomalies and predicts failures in real time to improve reliability and safety in critical operations.
LOCAL TRACK 03
Kaplan X
Salesforce
Learn More
Create an AI agent that proactively supports students in learning, wellbeing, and career readiness.
LOCAL TRACK 04
Lee Kong Chian School of Medicine
Learn More
Design an AI solution that supports people with dementia in Singapore through early detection, daily assistance, or predictive insights.
LOCAL TRACK 05
Micron
Learn More
Develop an LLM-powered smart parser that transforms diverse tool logs into clear, human-readable insights and structured database formats.
LOCAL TRACK 06
Singtel
Learn More
Build an adaptive AI system that detects, visualizes, and mitigates data drift to keep models accurate and reliable as data changes.
LOCAL TRACK 07
Workato
Learn More
Build a two-agent AI system that automates finance reconciliation and contract review while enabling both agents to collaborate seamlessly on business workflows.
Regional Track
AI Ready ASEAN
Learn More
Please refer to the following website for more information and to register: http://airayouthchallenge.ai
5 January ~ 16 February
Challenge Launch
Registration for all tracks will be open from
5 January (1200 SGT) to 16 February (1200 SGT)
. Please register early as registration may close once all slots are taken up.
Registration Close
Registration will close on
16 February (1200 SGT).
Successful applicants will be notified via email by 23 February.
16 February
2 March ~ 6 March
Meet the Partner & Challenge Announcement
Will be held from
2 to 6 March 2026
. Successful applicants will be invited to the session, where the challenge owner will provide an in-depth walkthrough of the problem statement and deliverables.
Submission & Finalist Selections
Details will be announced during the challenge announcement held in March.
April ~ May
22 May ~ 23 May
Grand Final @ AI Student Developer Conference
Finalists will be invited to present and compete in person at the AI Student Developer Conference on
22 & 23 May 2026.
Winners will receive their prizes during the award ceremony at the conference.
5 January ~ 16 February
Challenge Launch
Registration for all tracks will be open from
5 January (1200 SGT) to 16 February (1200 SGT)
. Please register early as registration may close once all slots are taken up.
16 February
Registration Close
Registration will close on
16 February (1200 SGT).
Successful applicants will be notified via email by 23 February.
2 March ~ 6 March
Meet the Partner & Challenge Announcement
Will be held from
2 to 6 March 2026
. Successful applicants will be invited to the session, where the challenge owner will provide an in-depth walkthrough of the problem statement and deliverables.
April ~ May
Submission & Finalist Selections
Details will be announced during the challenge announcement held in March.
22 May ~ 23 May
Grand Final @ AI Student Developer Conference
Finalists will be invited to present and compete in person at the AI Student Developer Conference on
22 & 23 May 2026.
Winners will receive their prizes during the award ceremony at the conference.
PAST NAISC GALLERY
LOCAL TRACK 01
LOCAL TRACK 01
Mission
Develop a Multimodal Security Response Advisor that helps officers assess incidents, recommend proportionate responses, and reinforce best practices using publicly available security guidelines and scenario-based examples.
Input Considerations:
CCTV footage and live camera feeds
Panic or distress calls from intercoms (e.g., lift breakdowns, accidents)
Access control logs and door alarms
Motion or intrusion sensor alerts
Note: Inputs are not limited to the above; teams may incorporate other relevant data sources or modalities.
Expected Output:
Context-aware and proportionate response recommendations aligned with standard security procedures and good practices.
Outputs can take various forms and are not limited to a text-based chatbot – participants are encouraged to explore diverse interaction modes or interfaces.
Any open-source resources or API to enable the development of this agent
Skills and Tools
Model Finetuning
Prompt Engineering
Python Programming
Deliverables
Slides on the developed agent, including design architecture, agent features, and future enhancements. There should also be a video demo recording of your agent on the slides.
Live demo of the developed agent [capped to 5 mins during NAISC Grand Finals]
Link to GitHub repository with README file
Eligibility
You must be a full-time student currently enrolled in a Polytechnic, University (Undergraduate) or postgraduate programme.
2 – 5 person(s) per team
Evaluation
To be confirmed
Prizes
Potential Internships Opportunities for Top 3 winning teams
Cash Prizes for Top 3 winning teams
Important Dates
Meet the partner session / Challenge Announcement – 6 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 02
LOCAL TRACK 02
Mission
Smart Multi-Agent System for Anomaly Detection and Predictive Maintenance
Critical infrastructure and industrial systems such as energy grids, water treatment plants, transportation networks, and manufacturing facilities—generate massive streams of operational and sensor data every second. Ensuring reliability, resilience, and security in these complex environments is challenging: traditional monitoring methods are largely reactive, addressing problems only after they occur, which can lead to costly downtime, safety risks, and operational inefficiencies.
Key challenges for AI solutions include:
Real-time anomaly detection: Identifying unusual patterns in sensor or operational data, including potential cyber-physical threats.
Predictive insights: Anticipating equipment degradation, system failures, or cascading impacts before they occur.
Decision support and explainability: Providing actionable, understandable insights to help human operators respond effectively.
Adaptive and collaborative intelligence: Leveraging AI approaches that can simulate, reason, or collaborate—such as agent-based methods—to monitor, predict, and act across interconnected systems.
Hackathon participants are encouraged to explore innovative AI solutions, including intelligent agents or multi-agent approaches, to create systems that can proactively detect issues, coordinate responses, and enhance overall operational awareness.
Skills and Tools
Agentic Solutions: Designing AI agents that act autonomously or semi-autonomously.
Situational Awareness: Building systems that monitor and interpret complex environments.
Human-in-the-Loop AI: AI agents assisting humans with root cause analysis and decision-making.
Intuitive UI/UX Navigation: Creating interfaces for clear visualization and interaction with AI insights.
Predictive & Anomaly Detection: Skills in forecasting, pattern recognition, and real-time monitoring.
Deliverables
One slide of the Multi-Agent AI System Solution, including
description of the design architecture.
Video demonstration of the Multi-Agent AI System built on Huawei Cloud.
Include text description that should explain the features and functionality of the project.
Text description should also include.
The relevant problem statement.
Development tools used to build the project.
APIs used in the project.
Assets used in the project, including models and datasets.
Libraries used in the project.
Include a link to the team’s public Github repository with README file. You could work towards rebuilding the results of your solution on benchmark.
Results and Performance of the Solution: Demonstrate the AI system’s performance on benchmark datasets/simulated scenarios, including metrics, visualizations, and effectiveness.
Eligibility
You must be a full-time student currently enrolled in Integrated Programme School (e.g. International Baccalaureate / NUS High Diploma), Junior College, Institute of Technical Education, Polytechnic, University (Undergrad), Post-Graduate or a Full-time National Serviceman (NSF)
2 – 4 person(s) per team
Evaluation
Novelty of the Solution
Practicality & Scalability
Technical Implementation
Presentation
Prizes
Potential Internships Opportunities for Top 3 winning teams
Huawei Products for top 1 team
Important Dates
Meet the partner session / Challenge Announcement – 4 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 03
LOCAL TRACK 03
Mission
Design an AI-powered agentic solution that helps students like you to organise their studies, stay motivated, and prepare for future careers.
Your solution should go beyond a chatbot answering questions. It should know when to reach out, what to recommend, and how to connect the dots across academics, wellbeing, and career readiness.
Choose one or more of these three strands to focus your solution:
Academic Success (Think: “You’ve missed two classes, want me to show you what to catch up?”)
Wellbeing & Motivation (Think: “You’ve got a major exam coming up soon. Would you like your school counsellor to share a few tips on managing stress?”)
Career & Future Readiness (Think: “There is a career roadshow on March 25th featuring fintech employers that match your interests. Should I register you?”)
Participants will use Salesforce Agentforce, a low-code agent builder — alongside Data Cloud and other Salesforce tools. Open-source technologies may also be incorporated.
The challenge:
How might you use AI and Salesforce tools to build an agentic solution that proactively supports students — helping them learn smarter, stay balanced, and prepare confidently for the future of work?
Skills and Tools
Salesforce Tools
Salesforce Agentforce (low-code agent builder to create digital assistants)
Einstein 1 Studio (prompts, skills, and actions)
Experience Cloud (optional portal UI)
Dashboards and Reports
Data Cloud (optional advanced insights)
Open-Source and External Tools (optional):
Bot frameworks such as Discord or Telegram
Programming languages such as Python or JavaScript
UI/UX tools such as Figma or Canva
Any open-source libraries for data, analytics, or integrations
No prior Salesforce experience is required; low-code options are supported. All registrants will be invited to attend preparatory workshop on Salesforce Low Code solutions.
Deliverables
One presentation slide overview of the proposed AI-powered solution (problem, users, value).
A 2-minute video demonstration showing the concept, prototype, or working dashboard/agent.
A short technical description explaining:
–
the problem addressed
– key features
–
solution architecture (Agentforce + optional open-source components)
Solution artefacts such as mockups, screenshots, flows, dashboards, or agent configurations.
Optional: A GitHub repository link with README if teams implemented code or bot logic.
Optional: User journey demonstration showing how a student would use the assistant
Eligibility
You must be a full-time student currently enrolled or waiting to start your studies in Upper Secondary, Junior College, Institute of Technical Education (ITE), Polytechnic, Private institution (Diploma), or a Full-time National Serviceman (NSF).
2 – 4 person(s) per team
Evaluation
Impact and relevance: How well the solution supports academic success, wellbeing, or career readiness.
Innovation and creativity: Novelty of the idea and meaningful use of AI and Agentforce.
User experience and design: Clarity, usability, and student-centred design.
Presentation quality: Clear explanation, strong storytelling, and good demonstration.
Prizes
Cash prize for top 2 winning teams
Cash Rebates for Kaplan Education programs for top 5 winning teams
Important Dates
Meet the partner session / Challenge Announcement – 4 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 04
LOCAL TRACK 04
Mission
As Singapore’s population ages, dementia has become a growing concern — affecting about 1 in 10 seniors aged 60 and above. People with dementia may struggle with memory, language, or daily activities, and this can be stressful for both them and their families. By 2030, the number of people living with dementia in Singapore could rise to more than 150,000, creating new challenges for healthcare, caregivers, and communities. While awareness is improving, early detection, daily support, and meaningful engagement for people with dementia still need innovative solutions.
The Challenge:
Your task is to design a creative AI-powered solution that can help detect, monitor, or support people living with dementia, their families, or caregivers. Think about how artificial intelligence — such as image recognition, natural language processing, chatbots, wearables, or smart environments — can be used to make life easier and more dignified for those affected.
The solution should show how AI can:
Identify patterns, risks, or emerging needs related to dementia, and
Create practical tools or interventions that improve safety, wellbeing, dignity, or caregiver support.
Solutions should be grounded in the Singapore context, considering an ageing population, family-based caregiving and multilingual communities.
Teams may choose to work with provided or open datasets to strengthen their solution.
Bonus points will be awarded to teams who mine datasets and incorporate predictive analytics to generate meaningful insights that directly inform or enhance their solution.
Goals:
Your project should show how technology can make a real and positive difference in dementia care. Consider how your idea could be used in Singapore — keeping in mind local lifestyles, multi-language needs, and the importance of compassion and respect for the elderly.
Impact:
By combining creativity, empathy, and technology, your solution could help families stay connected, caregivers feel supported, and seniors live with dignity and independence — contributing to a more caring, dementia-friendly Singapore.
Skills and Tools
All registrants will be invited to attend preparatory workshops on vibe coding and the topic of dementia.
Participants are encouraged to explore AI Singapore’s LearnAI platforms to gain a solid understanding of AI concepts and practical applications.
Relevant datasets related to the problem statement will be provided. Participants may also leverage publicly available open-source resources.
Deliverables
Team’s innovative approach to the problem statement / AI solution.
Challenges faced.
Reflections on the challenge.
Video demonstration of the idea and/or working prototype of the AI solution.
Eligibility
You must be a full-time student currently enrolled in Integrated Programme (e.g. International Baccalaureate/ NUS High Diploma) or Junior College
2 – 4 person(s) per team
Evaluation
Relevance of the solution.
Novelty of the solution.
Impact of the solution.
Feasibility of the solution (i.e. technical viability, practical applicability)
Presentation skills
Prizes
Certificate of participation for all teams
Top three winning teams will receive:
One trophy (Gold/Silver/Bronze) per team.
A medal for each team member.
Important Dates
Challenge Launch: 3 March 2026 @ LKCMedicine
Team Pitch Session: 12 May 2026 @ LKCMedicine
Final Presentation & Award Ceremony: 23 May 2026
Click here to register
LOCAL TRACK 05
LOCAL TRACK 05
Mission
Develop a Smart Tool Log Parsers with the use of LLM
Skills and Tools
Prompt Engineering
Python Programming Language
Relational Database
Any open-source parsers
Deliverables
Presentation slide with detailed illustration of design architecture and the flow of how the smart parser works on converting raw tool logs to human-readable contents and machine-readable table format in database (Log file ingestion, log file read and pattern recognition, tokenizing, normalizing, storage, query and analysis).
Handling of tool log input ranges from structured (Jason, XML, CSV), semi-structured (Syslog, Key-Value Pairs) and unstructured (plain text event log, non-human readable in binary or hexadecimal)
Text description that explains the features, functionalities, and constraints of the solutions:
Development tools used to build the project
APIs used in the project
Assets used in the project
A functioning prototype demonstrating capabilities using synthetic tool logs of different types
Eligibility
You must be a full-time student currently enrolled in a Polytechnic, University (Undergraduate) or postgraduate programme.
3 – 5 person(s) per team
Evaluation
Novelty of the solution
The variety of tool log file types the solution can handle and the comprehensiveness of different types of logs it can process
Presentation
Prizes
Certificate of participation for all teams
Top three winning teams will receive:
One trophy (Gold/Silver/Bronze) per team
A medal for each team member
Important Dates
Meet the partner session / Challenge Announcement – 5 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 06
LOCAL TRACK 06
Mission
Theme: Guarding Model Integrity in a Shifting Data World
Even the most advanced AI systems are only as reliable as the data that feeds them. When data quality slips or underlying patterns shift, AI decisions lose their grounding in reality. Detecting and addressing data drift is therefore not just a maintenance task—it’s a safeguard for trust and performance.
This challenge calls for solutions that uphold data integrity in a changing world, ensuring that AI remains accurate, resilient, and aligned with truth. Because if flawed data goes in, flawed intelligence comes out.
Track Challenge: Adaptive Drift Intelligence Challenge
Challenge Description
In this challenge, participants are tasked with building intelligent systems that can detect, visualize, and mitigate data drift — ensuring models remain robust as data distributions evolve.
Your solution should automatically identify and quantify data drift between a provided training dataset and a test dataset. It must be capable of analyzing a wide range of features, applying appropriate statistical and distance-based metrics to detect various types of drift, and highlighting features that exhibit significant changes.
For every identified drift, your system should include an intuitive interface that produces clear, insightful visualizations to help users understand the extent, direction, and nature of the shift.
Beyond detection, participants are required to design and implement automated mitigation techniques that adaptively handle the detected drifts during model training. These strategies should aim to preserve or enhance model performance under shifting data distributions.
Skills and Tools
Statistical Modelling
Model Fine-tuning and Training
Python Programming Language
Interactive dashboard development
*Datasets related to the problem statement will be provided.
*You are required to use open-source resources for the development of the AI Solution
Deliverables
Data drift detection framework that can:
Automatically identify feature types and apply the most appropriate statistical metrics to detect drift.
Adapt to new datasets with varying feature sets and data structures.
Generate intuitive visualizations that clearly illustrate the extent, nature, and potential impact of data drift for end users.
Documentation and presentation on the developed AI solution, including:
List of frameworks, and development tools used
Overview of the end-to-end solution workflow, including data drift detection and mitigation steps
Explanation and architecture of models used in the final solution
Detailed final model performance with relevant metrics
Justification for chosen approaches and comparison to other possible solutions
Limitations or known weaknesses of the solution
Frontend tool or dashboard (optional but recommended) to demonstrate data drift detection and correction in action
Public GitHub repository link with:
Structured project folders and clean codebase
README file
Dockerfile to enable reproducibility of solution
Eligibility
You must be a full-time student currently enrolled in Polytechnic, University (Undergraduate), postgraduate programme, or a Full-time National Serviceman (NSF)
2 – 5 person(s) per team
Evaluation
Participants will receive a dataset containing diverse feature types. Solutions will be evaluated based on:
Accuracy and completeness of drift detection
Effectiveness of mitigation methods in improving model robustness
Clarity and usability of the visualizations and interfaces provided
Performance and efficiency of the solution
Novelty of the solution
Scalability of the solution
Presentation
Prizes
Certificate of participation for all teams
Top three winning teams will receive:
1) One trophy (Gold/Silver/Bronze) per team.
2) A medal for each team member.
Potential internship opportunities for winning teams
Important Dates
Meet the partner session / Challenge Announcement – 6 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 07
LOCAL TRACK 07
Workato Challenge:
Mission
AI is improving quickly, but many enterprise AI projects fail to create real business value. This is not because the technology isn’t powerful, but because it lacks business context, proper controls, and safe ways to operate within enterprise systems.
MCP (Model Context Protocol) provides a standard way for AI to securely discover enterprise tools and data. But a
trusted, enterprise-grade MCP
can do so much more.
Your mission is to design and demonstrate an AI orchestration workflow powered by
MCP
. Build your solution to drive productivity and improve experience. Explain the potential business value in your submission.
What can you build?
The possibilities are endless. Here is a past award-winning entry that makes use of AI and could be reimagined with MCP:
First Place: (我)rk 爱 AI:
Designed to serve Singapore’s social impact sector, this solution helps resource-strapped charities by intelligently matching volunteers using an AI-assisted onboarding system.
This is just an example. The best solutions will surprise us – find a real problem, connect the right tools, and show what’s possible when AI is connected to applications – not just working in a silo.
Skills and Tools
What you’ll get:
A Workato Developer Sandbox account, complete with cutting-edge AI features including
Enterprise MCP
and
Agent Studio
Pre-built MCP servers to popular applications like Gmail, Notion, Asana, and more. Explore building your own custom integrations as well.
The freedom to define your own use case and demonstrate real business value to Workato and our community of path-defining customers and partners
Access to AI industry experts as your mentors
In-demand skills you will build:
AI Orchestration & Agentic Systems
: Design AI to pair with structured automation workflows that can be relied on to take action across business apps, not just generate text.
MCP (Model Context Protocol)
: Understand how MCP servers securely expose systems and data to AI models at enterprise scale.
Prompt Engineering for Action
: Craft effective prompts that guide AI to get the right context, tailor responses, take action, and handle edge cases reliably.
Systems Thinking & Workflow Design
: Map business problems to multi-step automated solutions. Think like a solutions architect.
Working with LLMs (e.g. Claude/OpenAI)
: Integrate large language models for summarisation, extraction, reasoning, decision-making and visualisation in real workflows.
Deliverables
A working prototype or concept demo that demonstrates how MCP can deliver real value.
A short demo video (2-4 mins) showing:
The problem
How MCP connects systems and controls data and actions securely
The impact
A presentation deck (max 8 slides) covering:
The workflow and integrations (what systems connect, what data flows where)
Key benefits and metrics
Potential impact
Evaluation
Innovation
– Creative, modern and agile approach to automating complex business processes with MCP
Use of Technology
– Effective use of MCP, LLMs (OpenAI/Claude) and Orchestration technology
Feasibility & Business Value
– Realistic, measurable impact for real-world users
Accuracy & Trust
– Reliable insights, clear explanations, and appropriate human oversight
User Experience
– How significantly does your quality of life improve?
Eligibility
You must be a full-time student currently enrolled in Integrated Programme (e.g. International Baccalaureate/ NUS High Diploma), Junior College, Institute of Technical Education (ITE), Polytechnic, University (Undergraduate), postgraduate programme, or a Full-time National Serviceman (NSF)
2 – 4 person(s) per team, ideally combining data, AI, and UX skills
Prizes
Certificate of participation for all teams.
Top three winning teams from Workato Challenge Track will receive:
One trophy (Gold/Silver/Bronze) per team
A medal for each team member
Cash prizes to be won
All registered teams will be awarded coaching hours with AI industry experts to provide guidance and support throughout the challenge
Finalists will gain access to exclusive mentorship and potential internship opportunities
Winners will earn special mention on Workato’s social media, win branded swag, and gain access to invite-only events with business and technology leaders, including Workato’s Senior Executives
Important Dates
Meet the partner session / Challenge Announcement – 5 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
REGIONAL TRACK
REGIONAL TRACK
Please refer to the following website for more information and to register:
http://airayouthchallenge.ai
Skip to content
REGISTRATION PERIOD
5 JAN
16 FEB
REGISTRATION
CLOSES IN
Days
Hours
Minutes
Seconds
About National AI
Student Challenge (NAISC)
Empowering students by providing them with the platform and resources needed to accelerate their AI learning journey, fostering innovation and cultivating the next generation of AI users, practitioners, and creators equipped to tackle real-world challenges.
Why Join?
Build Real-world AI Skills
Gain hands-on experience tackling authentic industry and societal challenges.
Win Attractive Prizes and Internship Opportunities
Top participants can earn attractive prizes and internships offered by track owners.
Boost Portfolio and Career Prospect
Strengthen applications for polytechnic, university, internships, and future jobs.
Learn from Industry Experts
Receive mentorship, attend workshops, and access guided resources from AI professionals.
The National AI Student Challenge 2026 offers 8 Challenge Tracks, each with its own eligibility criteria and requirements.
Challenge Tracks
STEP
01
Please carefully review the deliverables and eligibility criteria for each of the Challenge Tracks below.
STEP
02
Local Tracks are exclusively available to students studying in Singapore, while the AI Ready ASEAN Track is open to regional students studying
in Singapore, Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines, Thailand or Vietnam.
STEP
03
Each Challenge Track has limited space. Registration is on a first-come, first-served basis and will close once all slots are filled. Secure your spot by applying promptly. Successful applicants will be notified by
23 February 2026.
8 CHALLENGES
Discover more about the eligibility and criteria for each challenge!
LOCAL TRACK 01
Certis
Learn More
Develop a multimodal AI advisor that analyzes security inputs and recommends context-appropriate responses based on standard procedures.
LOCAL TRACK 02
Huawei
Learn More
Create a smart multi-agent AI system that detects anomalies and predicts failures in real time to improve reliability and safety in critical operations.
LOCAL TRACK 03
Kaplan X
Salesforce
Learn More
Create an AI agent that proactively supports students in learning, wellbeing, and career readiness.
LOCAL TRACK 04
Lee Kong Chian School of Medicine
Learn More
Design an AI solution that supports people with dementia in Singapore through early detection, daily assistance, or predictive insights.
LOCAL TRACK 05
Micron
Learn More
Develop an LLM-powered smart parser that transforms diverse tool logs into clear, human-readable insights and structured database formats.
LOCAL TRACK 06
Singtel
Learn More
Build an adaptive AI system that detects, visualizes, and mitigates data drift to keep models accurate and reliable as data changes.
LOCAL TRACK 07
Workato
Learn More
Build a two-agent AI system that automates finance reconciliation and contract review while enabling both agents to collaborate seamlessly on business workflows.
Regional Track
AI Ready ASEAN
Learn More
Please refer to the following website for more information and to register: http://airayouthchallenge.ai
5 January ~ 16 February
Challenge Launch
Registration for all tracks will be open from
5 January (1200 SGT) to 16 February (1200 SGT)
. Please register early as registration may close once all slots are taken up.
Registration Close
Registration will close on
16 February (1200 SGT).
Successful applicants will be notified via email by 23 February.
16 February
2 March ~ 6 March
Meet the Partner & Challenge Announcement
Will be held from
2 to 6 March 2026
. Successful applicants will be invited to the session, where the challenge owner will provide an in-depth walkthrough of the problem statement and deliverables.
Submission & Finalist Selections
Details will be announced during the challenge announcement held in March.
April ~ May
22 May ~ 23 May
Grand Final @ AI Student Developer Conference
Finalists will be invited to present and compete in person at the AI Student Developer Conference on
22 & 23 May 2026.
Winners will receive their prizes during the award ceremony at the conference.
5 January ~ 16 February
Challenge Launch
Registration for all tracks will be open from
5 January (1200 SGT) to 16 February (1200 SGT)
. Please register early as registration may close once all slots are taken up.
16 February
Registration Close
Registration will close on
16 February (1200 SGT).
Successful applicants will be notified via email by 23 February.
2 March ~ 6 March
Meet the Partner & Challenge Announcement
Will be held from
2 to 6 March 2026
. Successful applicants will be invited to the session, where the challenge owner will provide an in-depth walkthrough of the problem statement and deliverables.
April ~ May
Submission & Finalist Selections
Details will be announced during the challenge announcement held in March.
22 May ~ 23 May
Grand Final @ AI Student Developer Conference
Finalists will be invited to present and compete in person at the AI Student Developer Conference on
22 & 23 May 2026.
Winners will receive their prizes during the award ceremony at the conference.
PAST NAISC GALLERY
LOCAL TRACK 01
LOCAL TRACK 01
Mission
Develop a Multimodal Security Response Advisor that helps officers assess incidents, recommend proportionate responses, and reinforce best practices using publicly available security guidelines and scenario-based examples.
Input Considerations:
CCTV footage and live camera feeds
Panic or distress calls from intercoms (e.g., lift breakdowns, accidents)
Access control logs and door alarms
Motion or intrusion sensor alerts
Note: Inputs are not limited to the above; teams may incorporate other relevant data sources or modalities.
Expected Output:
Context-aware and proportionate response recommendations aligned with standard security procedures and good practices.
Outputs can take various forms and are not limited to a text-based chatbot – participants are encouraged to explore diverse interaction modes or interfaces.
Any open-source resources or API to enable the development of this agent
Skills and Tools
Model Finetuning
Prompt Engineering
Python Programming
Deliverables
Slides on the developed agent, including design architecture, agent features, and future enhancements. There should also be a video demo recording of your agent on the slides.
Live demo of the developed agent [capped to 5 mins during NAISC Grand Finals]
Link to GitHub repository with README file
Eligibility
You must be a full-time student currently enrolled in a Polytechnic, University (Undergraduate) or postgraduate programme.
2 – 5 person(s) per team
Evaluation
To be confirmed
Prizes
Potential Internships Opportunities for Top 3 winning teams
Cash Prizes for Top 3 winning teams
Important Dates
Meet the partner session / Challenge Announcement – 6 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 02
LOCAL TRACK 02
Mission
Smart Multi-Agent System for Anomaly Detection and Predictive Maintenance
Critical infrastructure and industrial systems such as energy grids, water treatment plants, transportation networks, and manufacturing facilities—generate massive streams of operational and sensor data every second. Ensuring reliability, resilience, and security in these complex environments is challenging: traditional monitoring methods are largely reactive, addressing problems only after they occur, which can lead to costly downtime, safety risks, and operational inefficiencies.
Key challenges for AI solutions include:
Real-time anomaly detection: Identifying unusual patterns in sensor or operational data, including potential cyber-physical threats.
Predictive insights: Anticipating equipment degradation, system failures, or cascading impacts before they occur.
Decision support and explainability: Providing actionable, understandable insights to help human operators respond effectively.
Adaptive and collaborative intelligence: Leveraging AI approaches that can simulate, reason, or collaborate—such as agent-based methods—to monitor, predict, and act across interconnected systems.
Hackathon participants are encouraged to explore innovative AI solutions, including intelligent agents or multi-agent approaches, to create systems that can proactively detect issues, coordinate responses, and enhance overall operational awareness.
Skills and Tools
Agentic Solutions: Designing AI agents that act autonomously or semi-autonomously.
Situational Awareness: Building systems that monitor and interpret complex environments.
Human-in-the-Loop AI: AI agents assisting humans with root cause analysis and decision-making.
Intuitive UI/UX Navigation: Creating interfaces for clear visualization and interaction with AI insights.
Predictive & Anomaly Detection: Skills in forecasting, pattern recognition, and real-time monitoring.
Deliverables
One slide of the Multi-Agent AI System Solution, including
description of the design architecture.
Video demonstration of the Multi-Agent AI System built on Huawei Cloud.
Include text description that should explain the features and functionality of the project.
Text description should also include.
The relevant problem statement.
Development tools used to build the project.
APIs used in the project.
Assets used in the project, including models and datasets.
Libraries used in the project.
Include a link to the team’s public Github repository with README file. You could work towards rebuilding the results of your solution on benchmark.
Results and Performance of the Solution: Demonstrate the AI system’s performance on benchmark datasets/simulated scenarios, including metrics, visualizations, and effectiveness.
Eligibility
You must be a full-time student currently enrolled in Integrated Programme School (e.g. International Baccalaureate / NUS High Diploma), Junior College, Institute of Technical Education, Polytechnic, University (Undergrad), Post-Graduate or a Full-time National Serviceman (NSF)
2 – 4 person(s) per team
Evaluation
Novelty of the Solution
Practicality & Scalability
Technical Implementation
Presentation
Prizes
Potential Internships Opportunities for Top 3 winning teams
Huawei Products for top 1 team
Important Dates
Meet the partner session / Challenge Announcement – 4 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 03
LOCAL TRACK 03
Mission
Design an AI-powered agentic solution that helps students like you to organise their studies, stay motivated, and prepare for future careers.
Your solution should go beyond a chatbot answering questions. It should know when to reach out, what to recommend, and how to connect the dots across academics, wellbeing, and career readiness.
Choose one or more of these three strands to focus your solution:
Academic Success (Think: “You’ve missed two classes, want me to show you what to catch up?”)
Wellbeing & Motivation (Think: “You’ve got a major exam coming up soon. Would you like your school counsellor to share a few tips on managing stress?”)
Career & Future Readiness (Think: “There is a career roadshow on March 25th featuring fintech employers that match your interests. Should I register you?”)
Participants will use Salesforce Agentforce, a low-code agent builder — alongside Data Cloud and other Salesforce tools. Open-source technologies may also be incorporated.
The challenge:
How might you use AI and Salesforce tools to build an agentic solution that proactively supports students — helping them learn smarter, stay balanced, and prepare confidently for the future of work?
Skills and Tools
Salesforce Tools
Salesforce Agentforce (low-code agent builder to create digital assistants)
Einstein 1 Studio (prompts, skills, and actions)
Experience Cloud (optional portal UI)
Dashboards and Reports
Data Cloud (optional advanced insights)
Open-Source and External Tools (optional):
Bot frameworks such as Discord or Telegram
Programming languages such as Python or JavaScript
UI/UX tools such as Figma or Canva
Any open-source libraries for data, analytics, or integrations
No prior Salesforce experience is required; low-code options are supported. All registrants will be invited to attend preparatory workshop on Salesforce Low Code solutions.
Deliverables
One presentation slide overview of the proposed AI-powered solution (problem, users, value).
A 2-minute video demonstration showing the concept, prototype, or working dashboard/agent.
A short technical description explaining:
–
the problem addressed
– key features
–
solution architecture (Agentforce + optional open-source components)
Solution artefacts such as mockups, screenshots, flows, dashboards, or agent configurations.
Optional: A GitHub repository link with README if teams implemented code or bot logic.
Optional: User journey demonstration showing how a student would use the assistant
Eligibility
You must be a full-time student currently enrolled or waiting to start your studies in Upper Secondary, Junior College, Institute of Technical Education (ITE), Polytechnic, Private institution (Diploma), or a Full-time National Serviceman (NSF).
2 – 4 person(s) per team
Evaluation
Impact and relevance: How well the solution supports academic success, wellbeing, or career readiness.
Innovation and creativity: Novelty of the idea and meaningful use of AI and Agentforce.
User experience and design: Clarity, usability, and student-centred design.
Presentation quality: Clear explanation, strong storytelling, and good demonstration.
Prizes
Cash prize for top 2 winning teams
Cash Rebates for Kaplan Education programs for top 5 winning teams
Important Dates
Meet the partner session / Challenge Announcement – 4 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 04
LOCAL TRACK 04
Mission
As Singapore’s population ages, dementia has become a growing concern — affecting about 1 in 10 seniors aged 60 and above. People with dementia may struggle with memory, language, or daily activities, and this can be stressful for both them and their families. By 2030, the number of people living with dementia in Singapore could rise to more than 150,000, creating new challenges for healthcare, caregivers, and communities. While awareness is improving, early detection, daily support, and meaningful engagement for people with dementia still need innovative solutions.
The Challenge:
Your task is to design a creative AI-powered solution that can help detect, monitor, or support people living with dementia, their families, or caregivers. Think about how artificial intelligence — such as image recognition, natural language processing, chatbots, wearables, or smart environments — can be used to make life easier and more dignified for those affected.
The solution should show how AI can:
Identify patterns, risks, or emerging needs related to dementia, and
Create practical tools or interventions that improve safety, wellbeing, dignity, or caregiver support.
Solutions should be grounded in the Singapore context, considering an ageing population, family-based caregiving and multilingual communities.
Teams may choose to work with provided or open datasets to strengthen their solution.
Bonus points will be awarded to teams who mine datasets and incorporate predictive analytics to generate meaningful insights that directly inform or enhance their solution.
Goals:
Your project should show how technology can make a real and positive difference in dementia care. Consider how your idea could be used in Singapore — keeping in mind local lifestyles, multi-language needs, and the importance of compassion and respect for the elderly.
Impact:
By combining creativity, empathy, and technology, your solution could help families stay connected, caregivers feel supported, and seniors live with dignity and independence — contributing to a more caring, dementia-friendly Singapore.
Skills and Tools
All registrants will be invited to attend preparatory workshops on vibe coding and the topic of dementia.
Participants are encouraged to explore AI Singapore’s LearnAI platforms to gain a solid understanding of AI concepts and practical applications.
Relevant datasets related to the problem statement will be provided. Participants may also leverage publicly available open-source resources.
Deliverables
Team’s innovative approach to the problem statement / AI solution.
Challenges faced.
Reflections on the challenge.
Video demonstration of the idea and/or working prototype of the AI solution.
Eligibility
You must be a full-time student currently enrolled in Integrated Programme (e.g. International Baccalaureate/ NUS High Diploma) or Junior College
2 – 4 person(s) per team
Evaluation
Relevance of the solution.
Novelty of the solution.
Impact of the solution.
Feasibility of the solution (i.e. technical viability, practical applicability)
Presentation skills
Prizes
Certificate of participation for all teams
Top three winning teams will receive:
One trophy (Gold/Silver/Bronze) per team.
A medal for each team member.
Important Dates
Challenge Launch: 3 March 2026 @ LKCMedicine
Team Pitch Session: 12 May 2026 @ LKCMedicine
Final Presentation & Award Ceremony: 23 May 2026
Click here to register
LOCAL TRACK 05
LOCAL TRACK 05
Mission
Develop a Smart Tool Log Parsers with the use of LLM
Skills and Tools
Prompt Engineering
Python Programming Language
Relational Database
Any open-source parsers
Deliverables
Presentation slide with detailed illustration of design architecture and the flow of how the smart parser works on converting raw tool logs to human-readable contents and machine-readable table format in database (Log file ingestion, log file read and pattern recognition, tokenizing, normalizing, storage, query and analysis).
Handling of tool log input ranges from structured (Jason, XML, CSV), semi-structured (Syslog, Key-Value Pairs) and unstructured (plain text event log, non-human readable in binary or hexadecimal)
Text description that explains the features, functionalities, and constraints of the solutions:
Development tools used to build the project
APIs used in the project
Assets used in the project
A functioning prototype demonstrating capabilities using synthetic tool logs of different types
Eligibility
You must be a full-time student currently enrolled in a Polytechnic, University (Undergraduate) or postgraduate programme.
3 – 5 person(s) per team
Evaluation
Novelty of the solution
The variety of tool log file types the solution can handle and the comprehensiveness of different types of logs it can process
Presentation
Prizes
Certificate of participation for all teams
Top three winning teams will receive:
One trophy (Gold/Silver/Bronze) per team
A medal for each team member
Important Dates
Meet the partner session / Challenge Announcement – 5 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 06
LOCAL TRACK 06
Mission
Theme: Guarding Model Integrity in a Shifting Data World
Even the most advanced AI systems are only as reliable as the data that feeds them. When data quality slips or underlying patterns shift, AI decisions lose their grounding in reality. Detecting and addressing data drift is therefore not just a maintenance task—it’s a safeguard for trust and performance.
This challenge calls for solutions that uphold data integrity in a changing world, ensuring that AI remains accurate, resilient, and aligned with truth. Because if flawed data goes in, flawed intelligence comes out.
Track Challenge: Adaptive Drift Intelligence Challenge
Challenge Description
In this challenge, participants are tasked with building intelligent systems that can detect, visualize, and mitigate data drift — ensuring models remain robust as data distributions evolve.
Your solution should automatically identify and quantify data drift between a provided training dataset and a test dataset. It must be capable of analyzing a wide range of features, applying appropriate statistical and distance-based metrics to detect various types of drift, and highlighting features that exhibit significant changes.
For every identified drift, your system should include an intuitive interface that produces clear, insightful visualizations to help users understand the extent, direction, and nature of the shift.
Beyond detection, participants are required to design and implement automated mitigation techniques that adaptively handle the detected drifts during model training. These strategies should aim to preserve or enhance model performance under shifting data distributions.
Skills and Tools
Statistical Modelling
Model Fine-tuning and Training
Python Programming Language
Interactive dashboard development
*Datasets related to the problem statement will be provided.
*You are required to use open-source resources for the development of the AI Solution
Deliverables
Data drift detection framework that can:
Automatically identify feature types and apply the most appropriate statistical metrics to detect drift.
Adapt to new datasets with varying feature sets and data structures.
Generate intuitive visualizations that clearly illustrate the extent, nature, and potential impact of data drift for end users.
Documentation and presentation on the developed AI solution, including:
List of frameworks, and development tools used
Overview of the end-to-end solution workflow, including data drift detection and mitigation steps
Explanation and architecture of models used in the final solution
Detailed final model performance with relevant metrics
Justification for chosen approaches and comparison to other possible solutions
Limitations or known weaknesses of the solution
Frontend tool or dashboard (optional but recommended) to demonstrate data drift detection and correction in action
Public GitHub repository link with:
Structured project folders and clean codebase
README file
Dockerfile to enable reproducibility of solution
Eligibility
You must be a full-time student currently enrolled in Polytechnic, University (Undergraduate), postgraduate programme, or a Full-time National Serviceman (NSF)
2 – 5 person(s) per team
Evaluation
Participants will receive a dataset containing diverse feature types. Solutions will be evaluated based on:
Accuracy and completeness of drift detection
Effectiveness of mitigation methods in improving model robustness
Clarity and usability of the visualizations and interfaces provided
Performance and efficiency of the solution
Novelty of the solution
Scalability of the solution
Presentation
Prizes
Certificate of participation for all teams
Top three winning teams will receive:
1) One trophy (Gold/Silver/Bronze) per team.
2) A medal for each team member.
Potential internship opportunities for winning teams
Important Dates
Meet the partner session / Challenge Announcement – 6 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
LOCAL TRACK 07
LOCAL TRACK 07
Workato Challenge:
Mission
AI is improving quickly, but many enterprise AI projects fail to create real business value. This is not because the technology isn’t powerful, but because it lacks business context, proper controls, and safe ways to operate within enterprise systems.
MCP (Model Context Protocol) provides a standard way for AI to securely discover enterprise tools and data. But a
trusted, enterprise-grade MCP
can do so much more.
Your mission is to design and demonstrate an AI orchestration workflow powered by
MCP
. Build your solution to drive productivity and improve experience. Explain the potential business value in your submission.
What can you build?
The possibilities are endless. Here is a past award-winning entry that makes use of AI and could be reimagined with MCP:
First Place: (我)rk 爱 AI:
Designed to serve Singapore’s social impact sector, this solution helps resource-strapped charities by intelligently matching volunteers using an AI-assisted onboarding system.
This is just an example. The best solutions will surprise us – find a real problem, connect the right tools, and show what’s possible when AI is connected to applications – not just working in a silo.
Skills and Tools
What you’ll get:
A Workato Developer Sandbox account, complete with cutting-edge AI features including
Enterprise MCP
and
Agent Studio
Pre-built MCP servers to popular applications like Gmail, Notion, Asana, and more. Explore building your own custom integrations as well.
The freedom to define your own use case and demonstrate real business value to Workato and our community of path-defining customers and partners
Access to AI industry experts as your mentors
In-demand skills you will build:
AI Orchestration & Agentic Systems
: Design AI to pair with structured automation workflows that can be relied on to take action across business apps, not just generate text.
MCP (Model Context Protocol)
: Understand how MCP servers securely expose systems and data to AI models at enterprise scale.
Prompt Engineering for Action
: Craft effective prompts that guide AI to get the right context, tailor responses, take action, and handle edge cases reliably.
Systems Thinking & Workflow Design
: Map business problems to multi-step automated solutions. Think like a solutions architect.
Working with LLMs (e.g. Claude/OpenAI)
: Integrate large language models for summarisation, extraction, reasoning, decision-making and visualisation in real workflows.
Deliverables
A working prototype or concept demo that demonstrates how MCP can deliver real value.
A short demo video (2-4 mins) showing:
The problem
How MCP connects systems and controls data and actions securely
The impact
A presentation deck (max 8 slides) covering:
The workflow and integrations (what systems connect, what data flows where)
Key benefits and metrics
Potential impact
Evaluation
Innovation
– Creative, modern and agile approach to automating complex business processes with MCP
Use of Technology
– Effective use of MCP, LLMs (OpenAI/Claude) and Orchestration technology
Feasibility & Business Value
– Realistic, measurable impact for real-world users
Accuracy & Trust
– Reliable insights, clear explanations, and appropriate human oversight
User Experience
– How significantly does your quality of life improve?
Eligibility
You must be a full-time student currently enrolled in Integrated Programme (e.g. International Baccalaureate/ NUS High Diploma), Junior College, Institute of Technical Education (ITE), Polytechnic, University (Undergraduate), postgraduate programme, or a Full-time National Serviceman (NSF)
2 – 4 person(s) per team, ideally combining data, AI, and UX skills
Prizes
Certificate of participation for all teams.
Top three winning teams from Workato Challenge Track will receive:
One trophy (Gold/Silver/Bronze) per team
A medal for each team member
Cash prizes to be won
All registered teams will be awarded coaching hours with AI industry experts to provide guidance and support throughout the challenge
Finalists will gain access to exclusive mentorship and potential internship opportunities
Winners will earn special mention on Workato’s social media, win branded swag, and gain access to invite-only events with business and technology leaders, including Workato’s Senior Executives
Important Dates
Meet the partner session / Challenge Announcement – 5 March
Final Presentation & Award Ceremony – 22 & 23 May
Click here to register
REGIONAL TRACK
REGIONAL TRACK
Please refer to the following website for more information and to register:
http://airayouthchallenge.ai