• Hackathon Date: April 25, 2026
• Submission Window: 10:00 AM – 10:00 PM (IST)

All submissions must be completed and uploaded on Devpost before the deadline. Late submissions will not be considered.

• The hackathon is open to students worldwide enrolled in a recognized educational institution.
• Participants may compete:
  • Individually, or
  • In teams of up to 4 members
• Cross-institutional and international teams are encouraged.
• Participants must register on Devpost to be eligible.

• All projects must be developed during the hackathon window (12 hours).
• Use of pre-trained models, libraries, and open-source tools is allowed.
• Participants must use the official datasets provided by the organizers.

Each submission must include:

• A clear project description
• A public GitHub repository with code and README
• Reported evaluation metrics (F1 Score / Accuracy / ROC-AUC as applicable)
• Screenshots or outputs demonstrating results

 Submissions must align with one of the official NLP challenge tracks.

AWS Skill Builder Annual Subscription + Oracle Cloud $300 Credit Account
Value: ~$749 USD + Microsoft 365 Personal (1-Year Subscription) Value: ~$70 USD

AWS Skill Builder Annual Subscription Value: ~$449 USD + Microsoft 365 Personal (1-Year Subscription) Value: ~$70 USD

🥉 Second Runner-Up (3rd Place)

Microsoft 365 Personal (1-Year Subscription)
Value: ~$70 USD

A powerful productivity suite for research, documentation, and development workflows.

CERTIFICATES

All registered participants who make a valid submission will receive a Certificate of Participation. Top 3 winners in each challenge track will receive a Certificate of Achievement

⚖️ Judging Criteria & Winner Selection

A panel of faculty and technical experts from the Department of Artificial Intelligence & Machine Learning, GGITS will evaluate all submissions.

Note: As there is no live presentation, evaluation will be based primarily on reported results, GitHub repositories, and documented outputs.

🧠 Evaluation Parameters (100 Points)

🔹 Model Performance & Accuracy — 40% ✅ (Primary Factor)

• Challenge 1 → Macro F1-Score
• Challenge 2 → Accuracy
• Challenge 3 → ROC-AUC

👉 Higher performance = higher score

🔹 Approach & Methodology — 20%

• Model selection and implementation
• Data preprocessing and feature engineering

🔹 Innovation & Creativity — 15%

• Unique ideas or improvements
• Advanced techniques or optimizations

🔹 Real-World Impact & Utility — 15%

• Practical relevance of the solution
• Scalability and usability

🔹 Documentation & Reproducibility — 10%

• Clear README and explanation
• Reproducible results
• Screenshots / outputs

🏆 Winner Selection

• Scores will be aggregated across all judges.
• The top three highest-scoring submissions, irrespective of challenge track, will be awarded:
  • 🥇 1st Place
  • 🥈 2nd Place
  • 🥉 3rd Place
• Winners will be officially announced within 72 hours after the datathon concludes.

 Important Rules

• Participants must clearly mention:
  • Evaluation method (train-test split / validation).
  • Dataset usage.
• Reported metrics must be verifiable.
• Any form of plagiarism or false reporting will lead to disqualification.