Aegython 2026

參加資格

學生限制

不限學生身分

關於這場比賽

Overview Organised by AeglysAI | Free Participation | Individual and Team Entries AEGYTHON 2026 is an AI/ML and Cybersecurity hackathon that challenges students, developers, engineers, and researchers to build practical, intelligent solutions for real-world problems. Participants will explore areas such as identity security, anomaly detection, API traffic monitoring, security alert explanation, outage prediction, and AI-powered solutions for everyday challenges. The focus is on working prototypes, technical understanding, responsible innovation, and reproducible results-not just ideas or presentations. Challenge Areas Participants can choose from six problem statements: Smart Login Guardian: Identify suspicious login attempts and explain potential risk signals. Find the Anomaly: Detect unusual patterns in synthetic data or system measurements. API Traffic Watch: Identify unusual API request patterns and traffic spikes using synthetic requests. Explain My Security Alert: Convert structured security alerts into clear explanations and safe recommendations. Predict the Outage: Analyse synthetic system metrics to identify early warning signs of possible outages. AI for Everyday Life: Build a practical AI/ML solution addressing a meaningful everyday problem. Eligibility Students, graduate students, early-career engineers, experienced practitioners, and researchers are the intended audience. Beginners and first-time contributors can explore the Foundation track; experienced participants may pursue more advanced engineering work. Participation is free. Every participant must complete the mandatory individual AeglysAI qualification and be selected before competing. Age and geographic eligibility must be confirmed in the final official event terms. Team Formation Participants may register individually. Teams may consist of 1-4 participants, including the team leader. Every team member must independently complete the qualification process and be selected. A team leader must submit the team roster for organiser verification and obtain a unique team ID. Participants cannot belong to multiple active teams. Team membership changes require organiser approval. Inter-college and inter-specialisation teams should be confirmed with the organisers before being advertised as permitted. Qualification Process Every participant must demonstrate basic familiarity with the AeglysAI project before selection. The qualification process involves: Registering individually through the official form. Forking the official AeglysAI repository into a personal GitHub account. Cloning the fork and demonstrating that the baseline builds or runs. Documenting the basic architecture and request/decision flow. Providing one meaningful technical observation, one realistic improvement idea, and a proposed validation approach. Submitting the required GitHub links and sanitised build evidence for organiser review. Completing qualification does not guarantee selection. Competition Format and Rounds Round 1 - Technical Screening Participants submit a working project, source code, a README with setup instructions, testing or validation evidence, and a demonstration video of no more than two minutes. Reviewers evaluate the solution's functionality, intelligent features, relevance, reproducibility, and video clarity. Round 2 - Finalist Demonstration Shortlisted finalists present their working projects live through Google Meet. Each finalist receives five minutes for the demonstration, followed by five minutes of technical questions and answers. Round 2 determines the final ranking. The two rounds are scored separately. Judging Criteria Round 1 - 100 points Working functionality: 30 points AI/ML or intelligent feature: 25 points Relevance and creativity: 20 points Reproducibility: 15 points Video clarity: 10 points Round 2 - 100 points Live demonstration: 35 points Technical understanding: 25 points Solution quality and safety: 20 points Technical Q&A: 15 points Presentation: 5 points Final Project Submission Requirements Each team must submit: Team ID and selected challenge Public GitHub source repository and a fixed review commit or tag Project README and reproducible setup instructions Demonstration video of up to two minutes Explanation of the AI/ML or intelligent technique used Testing or validation evidence Data and model attribution, licensing information, and known limitations A slide deck or walkthrough alone is not sufficient. The submission must be a working minimum viable product (MVP), executable prototype, or functional proof of concept. Any programming language or framework may be used. Competition Rules Participation is free; no registration fee is required. Every participant must complete individual qualification and selection before competing. Teams must follow the approved roster and team ID requirements.

主題與技術要求

資料來源