International Online Hackathon 2025 – Financial Analysis Agent with Explainable AI

As part of Super AI Engineer Season 5, under the collaboration between the Artificial Intelligence Association of Thailand (AIAT) and SCBX, we are excited to present the Explainable AI for Ethical Finance Hackathon 2025!

Hackathon Topic: Financial Analysis Agent

Objectives: Explore the future of AI-driven finance by developing Explainable AI-powered Ethical Financial Analysis Agents. Tackle real-world financial challenges through collaboration and experimentation, while prioritizing transparency and trust in AI models. Gain hands-on experience with open-source tools—especially Typhoon—and showcase their potential in creating responsible, practical financial solutions. This competition presents a unique opportunity to develop solutions that analyze complex financial data and promote trust and accountability in AI-powered finance.

Competition Focus: The participant will develop a financial analysis agent system utilizing large language models (LLMs) and agentic workflows designed to comprehend financial contexts, including investment strategies and economic principles, with responsibility for financial decision-making processes. These systems must demonstrate a comprehensive understanding of complex financial and economic scenarios while adhering to established ethical standards and cultivating user confidence in AI-driven financial analysis capabilities.

Competition Structure:

• Join Hackathon (Online)
• Final Pitching (Online)

Dates:

• Registration: 2-18 June 2025
• Problem Statement Release: 23 June 2025 (Online)
• Hackathon Duration: 23-27 June 2025
• Workshop Training: 23-24 June 2025 (Online)
• Pitching Day: 27 June 2025 (Online)

Hackathon Rules:

• Team size: Up to 15 members per team
• Open Source Models Only: Only open-source models are allowed. For example, Typhoon2 70B will be made available on the Lanta Platform.
• Dataset Usage: Participants may use any dataset, but must provide proper citations for all datasets.
• Basic System Prompts: Predefined system prompts will be provided, but participants can customize and improve these prompts to meet their specific solution needs.
• Evaluation Criteria: The AI models will be assessed based on weighted accuracy and speed. The model size will also be considered: Smaller models that deliver strong performance may earn extra points

🏆Show your skills and win big from a 5,000 USD prize pool!