02 / recognition
proof, not adjectives
Wins, rankings, and responsible-disclosure acknowledgements from global organizations.
// security acknowledgements
Responsibly disclosed vulnerabilities to global organizations: Lenovo, Panasonic, Oxford University, the Dutch National Cyber Security Centre (NCSC-NL), BuzzFeed, and multiple HackerOne programs.
12× National Hackathon Wins
12 national hackathon victories, including Bajaj HackRx 5.0 and Tech Future 2.0 at IIT Delhi.
Critical OAuth Auth-Bypass
Identified a critical OAuth authorization-bypass vulnerability in an internal enterprise application during the backend engineering role — a concrete web/API security finding inside production systems.
Published NLP Research — HCA-FND
Hybrid Context-Aware Fake News Detection (HCA-FND) — a published research paper. A two-stage NLP pipeline: TF-IDF + logistic regression for fast candidate filtering, then a DistilBERT transformer for deep semantic classification — balancing accuracy, latency, and interpretability for real-time use. Incorporates Named Entity Recognition and real-time fact-checking on extracted claims, reaching 96%+ accuracy across evaluation metrics.