Unleash Innovation:

Join the AI Hackathon 2025 Challenge! #AgentTime!

Welcome to the AI Hackathon 2025! This event is for Advantech colleagues passionate about AI technology. We will explore AI Agents’ possibilities, focusing on practical applications, encouraging innovative thinking, and discussing real-world challenges.

AI Agent

Unleash Innovation:

Join the AI Hackathon 2025 Challenge! #AgentTime!

Submission Requirements ( open to all Advantech employees) 競賽辦法與參賽須知

  • Hosted by the AI Orchestration Center, open to all Advantech colleagues ( including all RBUs ) to participate. Individuals or teams can participate, with a maximum of 6 people per team, to collaboratively create innovative AI Agents applications.
  • Complete the online form by the deadline and submit the original prototype or demonstration video (up to 10 minutes), proposal presentation, and other materials. Contest entries must not have won awards or been publicly presented elsewhere.
  • Any technical idea for creating "AI Agent" applications must tackle business pain points, quantify results like productivity gains or cost savings, and ensure the technology is feasible and proven effective.
  • The initial review will be conducted by business supervisors from various fields. Once approved, they will interact with the business supervisors and the AI technology team to optimize the works, and finally select A & A+ Cases in the finals.


  • Participants must complete the online submission form (form link is provided separately) and submit the following results before the submission deadline:
  • Work prototypes, demonstration videos or other content that can verify the effectiveness (conceptual proposals are not accepted)
  • The proposal brief or supporting documentation must clearly describe: application scenario, mplementation method, technical architecture, implementation results and future prospects.
  • The length of the achievement presentation video is limited to 10 minutes.
  • Note: The entries must not have won awards in other competitions, nor have they been publicly exhibited, published or shown.

So, What's an AI Agent ?

Think of AI Agents as intelligent helpers. They're computer programs that can learn from data, make choices, and take action to achieve specific goals, all without needing constant human guidance.

Example of a Non-AI Agent

Example of a AI Agent

Financial Multi-Agent:

• Problem Identification: Analyzes user questions to identify needs. 

• Data Retrieval: Extracts relevant financial data from databases. 

• Cross-System Integration: Connects data with other systems for collaboration. 

• Autonomous Response: Combines analysis to generate complete automated responses.

Example of a Non-AI Agent

Example of a AI Agent

NPI & RD Bug Smart Recommendation AI Agent:

  • Problem Analysis: The AI Agent analyzes departmental issues using RAG technology. 
  • Solution Search: It queries the internal problem database with the LangGraph model. 
  • Data Integration: Connects Microsoft Azure and Denodo databases with iDesign. 
  • Teams Access: Employees use Microsoft Teams and Agent Builder to access the AI Agent.


Example of a Non-AI Agent

Example of a Non-AI Agent:

General self-service Q&A Chatbot: It only queries fixed answers based on questions, lacking the ability to make proactive judgments and connect multi-step tasks.



What's This All About?

We're calling all you brilliant minds to team up, get creative, and build AI Agent solutions that'll totally transform how we do things and make our customers super happy.

Think of AI Agents as intelligent helpers. They're computer programs that can learn from data, make choices, and take action to achieve specific goals, all without needing constant human guidance.


Example of an AI Agent: Imagine a Financial Multi-Agent that grabs data, pulls info from databases, and connects with other Agent Groups to create a super-smart system.


Example of a Non-AI Agent: A Chatbot that just spits out answers. Helpful, but not an Agent in our book.

The mission: Utilize any technology to design and develop an application with "AI Agent" capabilities to accomplish at least one of the following objectives

競賽主題

Solve business process pain points, improve or enhance existing business processes, and expand business scope.

Clearly quantify the results, such as: productivity improvement, cost or time savings, creation of new value.

Utilize a feasible technical architecture with proven effectiveness.

The items apply to both the corporate proposal group and the free brainstorming group, with no difference in awards—only whether a submission topic is predetermined.

The Groups: You're free to select the group that best suits your vision and skills

活動組別

Free Idea Group

Participants choose their own case topics, from concept to implementation, and finally present and grade their results through demos. This category encourages creative thinking while focusing on technology implementation and application innovation.

Enterprise Challenge Group

The classic enterprise process framework (EPF) includes the value layer (creating customer value), the enablement layer (cross-departmental support), and the support layer (basic functions). Based on Advantech's characteristics, three key processes are selected from the EPF. Participants can choose one to design and implement an AI Agents solution to shorten the cycle and improve efficiency or satisfaction:

  1. TTM（Time to Market）
  2. OTD（On-Time Delivery）
  3. ITR（Issue to Resolution）

We Love Rewarding Innovation! 獎項與激勵

A+ Level Award

A+ Level Award

Award a certificate and a team bonus of NT$200,000 or a bonus of NT$30,000 per member

A Level Award & Special Contribution

A Level Award & Special Contribution

Award a certificate and a team bonus of NT$100,000 or a bonus of NT$20,000 per member

Promising Idea Award

Promising Idea Award

Whether or not the works passed the preliminary review, judges will consider providing incentives to teams or members of works that nearly won, based on their potential and development status.


•The works will be divided into five major areas: Sales & Marketing, R&D Engineering, SCM & Operation, Smart Manufacturing, and General Management (CoE). The number of awards may be vacant or adjusted upon decision by the jury in each field.

•The Special Contribution Award is for innovative applications that the jury identifies as having high potential commercial value or high productization potential.

•There is no difference in the awards for the Free Ideation Group and the Enterprise-assigned Group. The only difference is whether there is a pre-set submission theme.

Something for Your Inspiration 工具教學資源

Microsoft Copilot Studio

  • One-day tutorial: including use cases: ART_Copilot Studio In A Day_v2.pdf
  • Tutorial Video (3hr.): Provides a complete implementation guide for Copilot Studio
  • Microsoft ESI online teaching platform : provides advanced learning resources (must use Advantech email to register)

WISE AgentBuilder (Dify)

  • WISE-AI Agent/AgentBuilder_L1: Introductory course on basic Agent concepts and operations. 
  • WISE-AI Agent/AgentBuilder_L2: Advanced course on designing and deploying complex agents. 
  • New content released biweekly, featuring introductions and practical examples to aid quick mastery of tools.


Think of the hackathon as an AI playground! Team up with your crew, and build something epic. It's gonna be a blast, and you'll level up!