Synapse : The Kaggle Competition

Synapse is a high-octane, two-round machine learning competition designed to test your raw data endurance, feature engineering, and model optimization skills. Hosted at the Scaler School of Technology in Bangalore, this flagship event challenges data science enthusiasts to solve complex, real-world datasets under extreme time pressure.

• ₹30,000 prize pool, and top-tier networking opportunities.
• A rigorous two-round competition featuring an online qualifier (April 26–28, 2026) and a grueling 8-hour offline finale (May 17, 2026).
• Test your skills in exploratory data analysis, predictive modeling, and algorithm optimization.

For More Updates & Information stay tuned in the WhatsApp Channel

Round 1 (Online Qualifier):

• Register individually and form teams of 2 or 3 members.
• Gain access to a private Kaggle competition featuring a specially curated 3-day qualifier dataset.
• Build and train your models remotely. The Kaggle submission portal will automatically rank your team on the public leaderboard.
• Submissions can be updated multiple times until the deadline; your best score will be evaluated.
• Evaluation is strictly based on the designated scoring metric (e.g., RMSE, F1-Score) on the hidden test set.
• Only the Top 25 teams from the private leaderboard will earn a "Golden Ticket" to advance to the onsite Final Round.
  

Round 2 (The 8-Hour Offline Finals):

• A 8-hour in-person contest at Scaler School of Technology, Bangalore on May 17, 2026 (10:00 AM – 6:00 PM).
• Tackle a massive, complex, and uncleaned dataset revealed live at the venue.
• The Top 5 teams on the private leaderboard will face a rapid 10-minute code audit by the judging panel and deliver a 3-minute pitch on their model architecture.
• Lunch will be provided on-site to ensure the team can stay focused on the current training cycle.

Eligibility:

• Open to college students pursuing any tech or relevant program, data science majors, and AI/ML enthusiasts from anywhere in India.
• Basic familiarity with Python, Pandas, Scikit-Learn, and Kaggle submission formats is required.
• Participants from different colleges or institutions can form teams together. Cross-campus teams are fully allowed and encouraged.

Team Rules:

• Teams can have 2, or 3 members.
• Every participant must register individually first.
• Once all members have registered, the Team Leader forms the team by entering teammates' registered email IDs in the team formation form before Round 1 begins.
• Each registered participant receives an invite on their dashboard to join the team, which must be accepted to confirm.
• Once a team is confirmed, it cannot be changed; team assignments are permanent.

Rules & Regulations:

• Strict Compute Limits: To ensure a level playing field, teams may only use free-tier compute (Kaggle GPUs / Google Colab Free). Use of paid compute tiers (e.g., Colab Pro) will result in immediate disqualification.
• Participants must bring their own laptops and chargers.
• High-speed internet access will be provided at the venue.
• No external datasets or pre-trained models may be used unless explicitly permitted in the problem statement.
• All data cleaning, feature engineering, and model training must be executed during the competition hours.
• Participants qualifying for Round 2 must arrange their accommodation independently, as accommodation will not be provided.
• Finality of Decisions: The judges' decisions are final, binding, and not subject to appeal. No correspondence or discussion regarding the judging process or the final results will be entered into.
  

Registration Process:

• Step 1: Click on the “Register Now” button and complete your individual registration.
• Step 2: Create your team by adding members using their registered email IDs.
• Step 3: All invited members must accept the team invitation to confirm the team formation.
• Step 4: After completing the above steps, you will receive a confirmation email.
• Step 5: You will receive a further instructions along with the private Kaggle link for Round 1 on April 25, 2026.
• Step 6: Participate in the Round 1 online Kaggle contest, held from April 26 to April 28, 2026.
• Step 7: In 2-3 days, check your email to see if you have been selected for the finals or placed on the waitlist.
• Step 8: Arrive at the campus for the in-person Finals on May 17, 2026.
  
• Get event highlights, updates, and announcements on our  Instagram page.
  
• Note that it is mandatory for all participants to join the WhatsApp group, failure to do so will result in immediate disqualification.

Disqualification Criteria:

• Data Leakage (intentionally or accidentally training your model on the test dataset).
• Plagiarism, hardcoding outputs, or copying notebooks from external sources.
• Utilizing paid compute power or unauthorized external datasets.
• Misconduct or disruptive behavior during the onsite finals.
  

Participation Advisory:

• Laptop with a working charger.
• Any necessary local IDEs (Jupyter, VS Code) are pre-installed, though cloud notebooks are highly recommended.
• Note: Participants are welcome to explore nearby accommodation options or make stay arrangements based on their preference for the offline finals.
  
• Food and refreshments will be provided during the event.