KOJOH
Datasets nobody else's portfolio is built on.
Every dataset here starts as a real business problem, not a random CSV. Get the data, the story, and the challenge. Pick a folder below and see what's inside.
Browse by folder
Each folder is a job track. Open one to see what's inside.
UPI fraud detection, churn, quotation drop-off, and AI adoption ROI, the highest-demand track for job-seekers.
View datasets →Prediction done properly: honest labels, leak-free features, time-based tests and models worth rupees.
View datasets →Procurement, logistics and operations: transit trackers, cash plans and why things run late.
View datasets →Messy multi-source ETL, slowly-changing dimensions, and broken referential integrity to untangle.
View datasets →Ambiguous A/B tests, silent onboarding drop-off, and feature adoption metrics that mislead.
View datasets →Defect clustering across releases, coverage vs. incidents, and customer issue triage.
View datasets →