Sponsors
Data privacy and secure intelligence are no longer optional-they are the critical frontier of modern technology. Real innovation happens when organizations and individuals can leverage state-of-the-art agentic AI systems without risking sensitive data exposure. From healthcare records and proprietary research to financial systems and personal privacy, high-performance orchestration must happen locally and securely.
Building upon the success of our previous global competitions, the International Society of Data Scientists (ISODS) cordially invites high school students, university students, and industry professionals to the 6th World Championship in Data Science & AI (2026). Marking the milestone 10th Anniversary of ISODS, this year’s championship transitions into a premier, research-driven global conference format centered on "Privacy-First" Agentic AI with Yankee AI and Chroma DB.
The Generative AI World Championship centers on Yankee AI, a freemium, privacy-first agentic system developed by the George Washington Institute at ISODS. Built specifically for institutions, enterprises, and personal use, Yankee AI enables full Large Language Model (LLM) agentic workflows while maintaining complete local data security. Participating teams are challenged to engineer, optimize, and integrate a custom Chroma DB vector database to power secure, local retrieval-augmented workflows.
We challenge you to engineer solutions across key impact domains, including but not limited to:
Healthcare & Life Sciences.
Legal, Finance & Enterprise
Education & Academic Research
Public Sector & Global Resilience
Whether you are optimizing vector indexing strategies, evaluating embedding models, or tuning search performance, every contribution proves that high-performance agentic AI can be achieved without sensitive data ever leaving the local environment.
Your Goal
Tell a compelling story. Show the privacy-first problem, explain your Chroma DB engineering, and demonstrate how your integration with Yankee AI enables high-performance agentic orchestration locally. While technical execution is vital, your ability to communicate your scientific findings through a clear Writeup and video presentation will set the champions apart.
Submission Requirement
Submission Requirements
A valid submission must contain the following:
Kaggle Writeup
Attached Public Video (Presentation & Quick Demo)
Attached Public Code Repository
Attached Vector Database & Data Files (Shared Storage Link or Kaggle Dataset or Hugging Face)
Media Gallery
Your final Submission must be made prior to the deadline. Any un-submitted or draft Writeups by the deadline will not be considered by the Judges.
To create a new Writeup, click on the "New Writeup" button here. After you have saved your Writeup, you should see a "Submit" button in the top right corner.
Each team is limited to submitting only a single Writeup, but that same Writeup can be un-submitted, edited, and re-submitted as many times as you'd like. Your Writeup should contain a summary of your overall project along with links to supporting resources.
Note: If you attach a private Kaggle Resource to your public Kaggle Writeup, your private Resource will automatically be made public after the deadline.
Kaggle Writeup
The Proof of Work. This research-style paper report serves as your scientific and technical verification for the 6th World Championship in Data Science & AI (2026). It must clearly document the design and optimization of your custom vector database (using Chroma DB), how you seamlessly integrated it with Yankee AI, the challenges you overcame, and why your technical choices successfully support secure, local workflows. Its primary purpose is to prove to the international committee of professors and industry leaders that your presentation and demo are backed by rigorous research and real engineering.
The Kaggle Writeup serves as your research project report. This should include a title, subtitle, team details (1-3 members: high school students, university students, or professionals/alumni within 5 years of graduation), and a detailed technical analysis of your submission. You must select a Track for your Writeup in order to submit.
The below assets must be attached to the Writeup to be eligible.
a. Video (Presentation & Quick Demo)
Attach your video link to the Media Gallery. Videos must be 10 minutes or less and published to YouTube. This is a critical component of your submission. Create a dynamic, engaging, and high-quality video that covers both the technical background and a brief functional demonstration, structured into two distinct parts:
Part 1 - The Vector Database: Focus on the vector database itself. Explain your data sources, chunking strategy, embedding models, indexing, and performance optimizations within Chroma DB.
Part 2 - Yankee AI Integration & Quick Demo: Provide a walk-through and quick functional demo showing how your Chroma DB vector database integrates with Yankee AI to execute secure, privacy-first local workflows in action.
Your Goal: Tell a compelling story. Present the privacy-first problem, explain your Chroma DB engineering, and show a quick demo proving how your integration with Yankee AI enables high-performance agentic orchestration without sensitive data leaving the local environment.
Format: Post your video on YouTube and provide a direct link. It must be viewable by the judges without requiring a login.
b. Public Code Repository
The "Source of Truth". Provide a link to a public repository (e.g., GitHub, Kaggle Notebook). The code must be well-documented and clearly show the setup of Chroma DB, data ingestion, indexing, and vector database optimization. This is non-negotiable and will be used to validate the authenticity of your project. Your code repository should be publicly accessible and not require a login or paywall. If you use a private Kaggle Notebook, it will automatically be made public after the deadline.
You should provide a link to your public code in the section "Attachments" of your writeup, under "Project Links".
c. Vector Database & Original Data Files (Kaggle Dataset, Hugging Face, or Shared Storage Link)
A direct URL to your finalized vector database files along with your original data files and working materials. You may upload these assets directly as a Kaggle Dataset, a public Hugging Face Dataset/Repo, or provide a shared cloud storage link (e.g., Google Drive, OneDrive, Dropbox). This allows judges to evaluate your data engineering and perform qualitative verification firsthand. It must be publicly accessible and not require a login or paywall.
You can provide this link or attached dataset in the section "Attachments" of your writeup, under "Project Links" or "Datasets".
d. Media Gallery
This is where you should attach any images, architecture diagrams, benchmarks, research figures, and/or videos associated with your submission. A cover image is required to submit your Writeup.
Evaluation
Your project will be evaluated by an international committee of professors, researchers, and industry leaders through a two-stage review process: a preliminary Pass/Fail Audit for eligibility, followed by a Qualitative Scoring Review of your writeup, video presentation, code repository, and vector database assets.
Stage 1: Pass/Fail Eligibility Criteria
To advance to qualitative judging, submissions must satisfy all of the following baseline requirements:
Public Accessibility: The Code Repository, Vector Database/Data files, and YouTube video link must be publicly accessible without requiring logins, permission requests, or paywalls.
Submission Completeness: The submission contains all mandatory assets (Kaggle Writeup, YouTube Video link in Media Gallery, Code Repository link, Vector Database/Dataset link, and a Cover Image in the Media Gallery).
Video Format & Duration: The video is 10 minutes or less, published on YouTube, and clearly structured into Part 1 (Vector DB Architecture) and Part 2 (Yankee AI Integration & Quick Demo).
Code Authenticity: The public code repository contains working source code for Chroma DB setup, data ingestion, indexing, and vector DB optimization (not empty or template files).
Team Eligibility: The team consists of 1-3 members meeting the designated participant categories (high school students, university students, or professionals/alumni within 5 years of graduation).
Submissions that fail any single requirement above will be disqualified prior to qualitative scoring.
Stage 2: Qualitative Judging Criteria (100 Points Total)
Submissions that pass the eligibility audit will be evaluated by the jury based on the following criteria:
Top-performing teams will earn the championship's highest honors: international certification, widespread media coverage, and official peer-reviewed publication in a Special Issue of the Journal of Data Science and Artificial Intelligence (JDSAI). There will be a First Place, a Second Place, a Third Place, and three Honorable Mentions, tentatively.
Recognition
Recognition (Tentative)
International Recognition & Certification: The primary rewards for top-performing teams are official international championship certifications awarded by ISODS.
Peer-Reviewed Publication: Outstanding research writeups will be selected and invited for formal peer-reviewed publication in a Special Issue of the Journal of Data Science and Artificial Intelligence (JDSAI) celebrating the 10th Anniversary of ISODS.
Timeline
To accommodate academic calendars for high school and university students, align with global spring break schedules, and give participating teams sufficient time to finalize high-impact research deliverables, the final submission deadline for the 6th World Championship in Data Science & AI has been extended to the end of Q1 2026:
March 31, 2026 (End of Q1) - Extended Final Submission Deadline (11:59 PM UTC).
April 2026 - International Jury Evaluation & Announcement of Championship Finalists.
May 2026 - JDSAI Special Issue Publication.
All deadlines are at 11:59 PM UTC on the corresponding day. The competition organizers reserve the right to update the contest timeline if deemed necessary.
Sponsorship
Sponsorship
We extend our sincere appreciation to Google for their sponsorship of Google Ads.
Committee
