Deeksha Karthik, a 12th-grade student at Oberoi International School, has spent the past two years working on projects about online financial fraud in India. She began her journey with awareness campaigns, then did academic research and gained industry experience. Now, she has created a cybersecurity product being used across 3 countries.
Deeksha’s interest in helping people avoid financial fraud initially developed at the foundational level. Since August’23, she has co-founded Project MoneySatark, an initiative that has helped raise awareness through a website. It organized interactive community meetings and held more than ten sessions across networks. It later secured funding from the IB Global Youth Action Fund. Eventually, this work led her to engage directly with people about how cameras would affect them, so she took it upon herself to understand the gap and set out to fill it.
Her early insights led to academic research and a paper called “Proactive Cybersecurity Using Machine Learning: Threat Detection, Diagnosis, and Propagation Analysis,” co-written with Anand Zutshi and published in the International Journal of Science and Research (IJSR) in September 2025. The study
explores how text analysis, graph examination modeling, and machine learning can improve cyber detection and handling. This research became a key foundation for her later projects.
She also interned at Qdit Labs, a company working on quantum-safe cybersecurity solutions for businesses and financial systems. This experience gave her a firsthand look at how cybersecurity is handled in industry, and she used what she learned in her next project.
Deeksha brought together community insights, research, and industry experience to create Kedayam, a free and open-source Google Chrome extension. It protects everyday internet users in real time from phishing attacks, ClickFix malware, data leaks, and tech-support scams, all without sending any user data to the cloud.
Online financial fraud is one of the fastest-growing crimes in India. In 2025, this type of cybercrime caused losses of over ₹981 crore. Most people do not have good ways to protect themselves from the browser-based methods behind these crimes. Kedayam runs quietly in the background and warns users before they click, enter a password, or share financial information.
The extension’s detection engine works entirely on the user’s device. It uses a logistic regression classifier trained on 12,000 URLs, along with heuristic and rule-based analysis of page content, form behavior, and authentication processes to spot threatsthatsignature-based tools often miss. Importantly, the system does not use cloud processing, does not require an account, and collects no telemetry, so users are protected without sharing any personal information.
Kedayam has passed 14 rounds of security audits and is supported by 695 automated tests across 82 test files. It currently has a detection precision of 97.98% and a false positive rate of only 0.8%, which is similar to many larger commercial security products. Since its launch, more than 50 users in India, the United States, and the United Kingdom have downloaded it, and it has already blocked over 280 threats in real-world situations.
Deeksha continues to work on Kedayam, aiming to improve its threat detection and reach more users. Her journey from community outreach to publishing research and building a fully functional, independently verified security product is a rare example of combining academic rigor, technical skill, and dedication to solving a real problem.