Datanyx'26

Overview DATANYX'26 is Telangana’s flagship 24-hour datathon, now in its third edition. It is organized by AWS Student Builder Group MJCET and IEEE SMC MJCET. It is a 24-hour event where you take raw data and turn it into a working prototype. You start from scratch. The problem statements, datasets are given to you on the first day. DATANYX started in 2024 as the first 24 hr datathon in Telangana. In 2025, it grew into a national event. This year, every problem is linked to the UN Sustainable Development Goals. So what you build can help solve real problems in the world. Domains: AgroTech (SDG 2, 12, 13, 15) EdTech (SDG 1, 4, 8) Urban Tech (SDG 6, 7, 9, 11) MedTech (SDG 3, 5, 10) Event Details: Dates: 28th and 29th November 2026 Venue: Ghulam Ahmed Hall, MJCET, Hyderabad Mode: Offline Guidelines: Eligibility: All undergraduate students Team size: 2 to 4 members Inter-college team members: Allowed Duration: 24 hours Fields of Study: Teams can be formed from various specializations, including AI&DS, AI&ML, CSE, IT, ECE, etc. Event Structure: Phase 1 - Shortlisting Round (Online): In the first phase, teams are required to submit a unique problem statement of their own creation, accompanied by a solution abstract with a presentation that aligns with their selected domain. Teams must choose from one of the following domains: Med Tech, Urban Tech, Agro Tech or EdTech. The proposed presentation should reflect originality and innovation within the chosen domain.All submissions must be completed online on the provided ppt template before the designated deadline. The DATANYX'25 PPT TEMPLATE & THE DOMAIN DESCRIPTION are attached below . Phase 2 - Finals (MJCET): The top 15 teams from each domain will qualify for the final phase, held in person at MJCET, Hyderabad. Here, teams will be given a new problem statement revealed on the spot along with a sample dataset. They will then have 24 hours to brainstorm, design, implement, build, and submit their final solutions. This phase tests the teams' ability to think critically and collaborate effectively under pressure, turning ideas into impactful solutions within a limited timeframe. Shortlisted teams are required to pay a participation fee of ₹800. Step 1: Getting Started: Click the Unstop link for Datanyx. Find the "Register" button and click it. Step 2: Sign In (Super Easy!): Choose your sign-in option: Google (fastest!) LinkedIn Email Enter your details: Phone number Email Wait for the OTP on your phone and enter it. Step 3: Your Basic Info: Fill in the following details: Your name School or workplace Course details Graduation date Country of residence Step 4: Team Time: If you’re the Team Leader: Create your team: Choose a team name. Add teammates' emails. Remind them to check their email for the "View Opportunity" link. Watch for teammates' names to turn GREEN once they join! If you're joining someone’s team: Find the team name. Click "Send Request". Wait for the team leader to add you. Step 5: PPT Submission: Download the PPT template Create your PPT with the team Submit your PPT in the PPT  Submission box. Step 6: Final Check: Find the "View Details" button. If it's GREEN = You’re all set! If it's RED = Someone still needs to confirm. Pro Tip: Make sure everyone confirms their participation—green names mean your team is ready! Rules: Dataset Verification: Teams must use a real, existing dataset. Register it within the first 2 hours with the source link, references, and a file hash. The same check is repeated at final submission, so the data cannot be changed midway. Live Injects: The problem does not stay fixed. Every 4 to 6 hours, a new inject drops, such as a new feature, a changed direction, or new data. Judging covers robustness and explainability, not only accuracy. Commit Trail: Commit your work at least every 90 minutes with meaningful commit messages. Judges review the history. A large block of finished code appearing at once will be questioned. Fresh Start Rule: All work must start after the event begins. Organizers create each team’s repository at the opening. Only open-source libraries are allowed. Pre-written or pre-built projects are not allowed. AI Use Log: Keep a short log of what you asked AI and what you used. Undisclosed AI use is penalized. All AI is banned during the Dark Hour, an unannounced one-hour window in the second half of the event. Real User Validation: After building the product, at least 15 real users who are not on your team must test it. Testers fill in a short form on whether they understood it, whether they would use it, and whether it solves the problem. Organizers will contact a few testers at random to verify. API Transparency: Declare every API and external service you use. Calls found in the repository scan but missing from your declaration affect your score. Static or hardcoded responses shown as live outputs will be flagged. Business Approach: Submit a one-page business plan covering the customer, who pays, how the product earns or saves money, running costs, competitors, and how you will reach your first 100 users. A judge with industry experience will question your numbers at the defense. You are scored on how realistic the plan is, not on how big the claims are. Public Notebook Plagiarism Check: Submissions are checked against public Kaggle notebooks and GitHub repos that use the same dataset. Reusing public code is allowed only if it is clearly cited. Uncited copied code is penalized. Pre-trained Model Declaration: You may use pre-trained models from Hugging Face or other sources, but you must declare the base model and show your own training or fine-tuning logs. Downloaded fine-tuned weights presented as your own work count as a violation. Outside Help Rule: Only registered team members can contribute to the project. Mentors can guide teams but cannot write code for them. Commits from unregistered GitHub accounts lead to disqualification. The organizers’ decision is final. Judging: Teams are judged on accuracy, robustness, explainability, and how realistic their business plan is. Prizes: Rs 1,00,000 prize pool.