

.
During DMRC internship in Mechanical Section, specialized in bogie suspension systems, fire fighting, brake/pneumatic systems, and metro door operations. Gained hands-on expertise in HVAC systems critical for passenger comfort and safety. Contributed to locomotive engine refinement projects enhancing efficiency. Developed strong foundation in metro rail mechanical engineering systems.
• Developed comprehensive HVAC data analytics solution processing [X] data points
daily from building automation systems, implementing seasonal decomposition
algorithms (STL/ARIMA) for trend analysis and isolation forest/z-score methods
for outlier detection,
•enabling data-driven optimization of chiller operations,
AHU scheduling, and thermal comfort.
Built end-to-end data pipeline for airport FFS/FIDS analytics, integrating
real-time flight data.
•Anomaly detection on RFID/sensor streams flags jams 30min early (features: velocity variance, queue length). Achieves 95% precision using XGboost / Light GBM
•predict BHS jams (jam/no-jam) from conveyor metrics (load, speed variance), or PBB faults (aligned/misaligned) via dock sensors. Coefficients reveal key drivers like peak-hour volume; regularized L2 variants handle multicollinearity in high-dimensional airport IoT streams.
Analytical Framework : Pandas , Numpy , Matplotlib , Sckitlearn Power BI ,XGboost and pytorch
1 ) Clinical/ Pharma Data analytics Dashboard. 30-07-2025
•Built an analytics solutions for real time clinical trial dashboard for Salesforce optimization helps in integration of external and internal data
2) Pharma copay fraud detection ( Supervised ML ) (25-09-2025)
•Built an innovative analytical solution for the anomaly and error detection in real world pharmacy
•False positives by 20-35 percent calibrated thresholds
3) Built Churn propensity Model using LR and KNN
• Delivered ROC-AUC-0.84 Life@10%+28% and cut churn by targeted retention offer ; Deployed as a Rest API with daily batch scoring