Potsdam, New York, United States
Currently a PhD Research Assistant in Civil and Environmental Engineering at Clarkson University, with research focused on concrete, alkali-activated systems, hybrid cementitious materials, and integrating machine learning techniques for materials characterization and performance prediction.
Worked on advanced AI solutions including global forecasting systems (24% MAPE across 300+ categories using XGBoost, LSTM, Prophet), SHM and Damage Detection of Concrete using vision models, agentic LLM platforms using Google A2A & Anthropic MCP, and ETL pipelines with Airflow and AWS. Built signature verification (82.4% accuracy with Siamese ResNet), molecular generation models (AutoEncoders), and a dynamic forecasting platform supporting multiple models. Also developed an LLM-powered pharma assistant using LangChain, RAG, and knowledge graphs.
Built ML models for brake wear prediction using telemetry data with 92% accuracy (XGBoost) and engineered features like Brake Stress Index to optimize servicing. Processed vibration sensor data for early engine fault detection, improving diagnostic accuracy by 15%. Focused on feature engineering, sensor data analytics, and predictive maintenance in automotive. Expertise Pedal Deflection analysis and Modal analysis using Finite Element Analysis
Reviewed AutoCAD drawings, Analysis and Design of G+3 Residential building using STAAD Pro software, Analysis and Design of steel portal frame.