# GCC 基金公共資助申請/申請贈款 活動類

:::success
里程碑1,2
:::
## 里程碑1目標詳述
Our main goal during the Kaggle competition is to ensure smooth operations while addressing technical issues in real-time and gathering participant feedback for future improvements.
## 里程碑1資金使用計劃

- Deliverables
* Gather feedback on the dataset and task design for future improvements
* Produce post-competition analysis and suggestions
* Summarize errors and provide optimization recommendations for the next phase

- **TOTAL : 8,000**

| Task | Description | Budget |
| -------- |-------- |-------- |
| Kaggle Prize Pool | Host a reward-based competition on Kaggle, aiming to attract more AI talent to participate. | 5000 |
|Technical Support During Competition|Provide technical support throughout the entire competition period, including responding to participants' questions regarding the dataset, task definitions, formats, and scoring processes.|1,000|
|Submission & Scoring Troubleshooting|Assist in resolving scoring errors, submission anomalies, and other technical issues that affect the competition process.|1,000|
|Evaluation System Monitoring|Monitor the consistency of public/private evaluations and address edge cases and submission interruption issues.|1,000|

## 里程碑2目標詳述
- Goal:
Verify and reproduce submitted solutions and participation, gather feedback, and document insights to improve future competitions and ensure dataset integrity.
## 里程碑2資金使用計劃

- Deliverables:
* Prepare articles or video tutorials to share the experimental results with the community.
* Share insights, potential future improvements, and the performance of LLM in identifying the given vulnerability with more developers and security researchers in the industry. Invite AI experts to join and contribute to the effort.

- **TOTAL : 4,000**

| Task | Description | Budget |
| -------- |-------- |-------- |
|Solution Execution Costs | Fees for running Kaggle competition submitted solutions| 2,000 |
|Solution Execution & Verification|Run the winning solutions on the full evaluation dataset to verify their performance.Reproduce submitted results to confirm correctness and consistency with leaderboard scores.Identify any anomalies or errors in submitted solutions.|1,000|
|Participant & Submission Summary|Compile a comprehensive list of all participants, teams, and submission counts.Identify high-performing participants and notable solution approaches.|500|
|Post-Competition Feedback Collection|Collect feedback from participants regarding dataset clarity, task definitions, and submission process.Identify pain points or ambiguities in the competition workflow.Summarize participant suggestions for improving future competitions.|500|

## Team Summary
### Daky Wang

Daky is the Product Development Lead at OneSavie Lab, a seasoned professional with over 15 years of experience in software engineering. In the Bastet project, Daky is responsible for initial idea brainstorming and market research from scratch, and during the project, he handles system architecture planning and requirement definition.
In the past 3+ years, we have built

* LazyOtter Finance:
A scroll-based DeFi project that assesses risk profiles of decentralized finance products. It provides users with risk-adjusted yield optimization and real-time on-chain monitoring to identify and respond to security threats swiftly. In terms of industry recognition, its project LazyOtter Finance achieved 2nd place at both ETH HongKong and Bybit Demo Day, and was one of the top 10 projects on Scroll Chain in terms of TVL.
website: https://lazyotter.finance

* Mamori AI:
Delivers AI-driven risk intelligence solutions for the Web3 sector. This platform integrates advanced analytics, live monitoring, and off-chain data processing to detect threats and simplify risk management.
website: https://mamori.ai

Public speaking experience:
* CyberSec 2025
https://cybersec.ithome.com.tw/2025/speaker-page/1723
* Eth Taipei 2025
https://ethtaipei.org/agenda#info

Hackathon experience:
* 2023 TBW hackathon award $500: Learnverse
ref: https://x.com/BNBCHAIN/status/1738152023361511727
* 2025 ETH Global Taipei Hackathon: Messiah
HashKey – On-Chain Infrastructure Compliance Innovation
Celo – L2 Pool Prize
Here’s the link to our award-winning project:
ref: https://ethglobal.com/showcase/messiah-d6zo2

Additional Information about Daky :
LinkedIn: https://www.linkedin.com/in/dakywang/

### Alice Hsu
Alice Hsu is a researcher focused on Web3 security.

She has worked in the Metaverse department at Trend Micro developing vulnerability detection tools that combine AI and static analyzers.
She serves as an auditor and judge on several competitive audit platforms.
Currently, she is a security researcher and dedicated to promoting security knowledge, developing tools, and conducting research.

She has shared her research at conferences such as CyberSec 2022, CyberSec 2024, CyberSec 2025 , and Eth Taipei.

She is also part of the operations team and a white hat member in the DeFiHackLabs community.

She was an instructor at the Web3 Security BootCamp sponsored by the Ethereum Foundation.

In the Bastet open source project, she is responsible for leading the primary vulnerability labeling efforts and collaborating with community white hats to ensure the quality of the vulnerability labels. She also works with AI experts to integrate past audit experience and knowledge into the vulnerability detection workflows.

Additional Information about Alice Hsu :
X : https://x.com/AliceHsu_kou
Public speaking experience:
- CyberSec 2022
https://cybersec.ithome.com.tw/2022/speaker-page/492
- CyberSec 2024
https://cybersec.ithome.com.tw/2024/speaker-page/492
- CyberSec 2025
https://cybersec.ithome.com.tw/2025/speaker-page/492
- Eth Taipei 2025
https://ethtaipei.org/agenda#info

### Kevin Ke

Kevin Ke is an AI researcher specializing in AI security.
During his master's studies, his research focused on using deep learning, specifically LSTM methods to analyze logs and predict the likelihood of potential issues.

In the past few years, he has served as an AI & cybersecurity consultant, assisting multiple companies in the IC design, semiconductor, finance, and insurance industries.

He is currently an AI security researcher at AIFT, dedicated to promoting AI security knowledge, uncovering new risks, developing GenAI vulnerability scanning and protection tools, and conducting research.

In the Bastet project, he leverages his past experience to improve experimental processes and participates in the design and planning of benchmarks.

Public speaking experience:
- informationsecurity , 2025
https://www.informationsecurity.com.tw/article/article_detail.aspx?aid=11933

- ICIM 32th , 2021
https://sites.google.com/view/icim2021/%E6%9C%83%E8%AD%B0%E8%AD%B0%E7%A8%8B?authuser=0

Additional Information about Kevin :
LinkedIn : https://www.linkedin.com/in/kevink-trk/

### Chengyu
Chengyu's expertise lies in machine learning and Web 2.0 backend development, with experience in developing static analysis tools and data preprocessing. By designing and implementing processes that transform code into structured features, he helps improve the efficiency and recognition capabilities of AI models and detection engines. He is a key player in the Bastet open source project, participating in dataset format planning and collaborating with security researchers to develop AI vulnerability detection processes and optimize LLM performance.