Ritwick Mondal

I am a Computer Science and Engineering undergraduate with practical experience building ML and NLP systems for cybersecurity and content moderation. I combine strong CS fundamentals with hands-on experimentation across datasets, tooling, and model evaluation.

During an internship at DRDO (CAIR Lab) I designed a full ML pipeline for encrypted network-traffic classification, applied domain-informed feature selection, and achieved 99.3% accuracy with LightGBM using seven features. I evaluated nine algorithms across multiple performance metrics and worked with flow-extraction tools like Tranalyzer2 and Zeek.

My projects include ToxicCommentNet, a multi-label NLP classifier using DNN/BiLSTM/BiGRU (93.89% best accuracy); MalScan, a large-scale lexical detector for suspicious URLs built from millions of samples; and ForecastFly, a time-series forecasting project comparing SARIMA, Auto ARIMA, Prophet, and Holt-Winters models. I emphasize rigorous preprocessing, evaluation, and visualization.

I contribute in team settings as a senior member of Robocell CCA, leading workshops and coordinating events. I seek roles where I can apply ML/NLP and cybersecurity skills to build production-ready models and systems that deliver measurable impact.