Vineeth

Experience

1. Data Scientist

   Cequence Security

   Dec 2018 — Present

   I develop applications that utilize AI/Machine-Learning for novel threat detection. I\'m responsible for designing automation systems to mitigate risk, fraud, and protect sensitive web assets• Owned projects end-to-end, from problem scoping to ML model deployment and metriccommunication to stakeholders.• Enabled developers to build secure AI agents that can consume enterprise APIs over MCPprotocol• Enhanced browser forgery detection with an ensemble decision tree model • Analyzed multiple timeseries for periodicity using Deep Learning model with an AUC of 0.7• Built a lightweight classifier in conjunction with feature vector similarity to differentiate goodand bad fingerprints, achieving an F-1 score of 0.7 and a maximum False Positive Rate of 2%• Delivered effective insights with statistical A/B tests using a champion-challenger model• Reduced number of threat incident alerts and related on-call man-hours by more than 25%• Leveraged ELK stack’s Filebeat for Airflow logging to reduce customer friction by 15% innew Cequence Kubernetes app installations• Developed ETL pipelines to train, on a schedule, multiple forecasting (fbProphet) models that predict browsers request rate with an average SMAPE score of 25%• Enabled junior data scientists to take on more complex projects through continuous support