• Led a team of 5 software developers and 7 data analysts to improve existing ETL pipelines• Designed and developed algorithm to process IMS‐MS multiplexed data using CNNs‐Patent in Progress• Worked on a algorithm to find Centroid of LC‐MS and IMS‐MS data‐(Patented Algorithm)• Formulated an algorithm to accurately model non‐linear Retention Time shifts.• Hosted the LC‐MS data processing software on AWS for public usage• Performed extensive Multivariate Statistical Analysis… Show more • Led a team of 5 software developers and 7 data analysts to improve existing ETL pipelines• Designed and developed algorithm to process IMS‐MS multiplexed data using CNNs‐Patent in Progress• Worked on a algorithm to find Centroid of LC‐MS and IMS‐MS data‐(Patented Algorithm)• Formulated an algorithm to accurately model non‐linear Retention Time shifts.• Hosted the LC‐MS data processing software on AWS for public usage• Performed extensive Multivariate Statistical Analysis to obtain Significant Metabolites• Achieved accuracy of 95% on multiple statistical model for heart health, diabetic health and kidney health score for Human Plasma samples.• Performed Dynamic Flux Balance Analysis on CHO Genome-Scale Metabolic Model• Streamlined the Genome-Scale Metabolic Pipelines using Python to eradicate human expert time.• Streamlined the data transfer using SSH, designed and developed software for LC‐MS data glacier management• Led and managed 5+ client projects simultaneously to meet the strict deadlines Show less

• Designed and built ETL data pipelines to process LC‑MS, IMS‑MS and tandem MS/MS data• Used Builder design pattern and SOLID principles to design the ETL data pipeline• Developed base module that can be used in multiple ETL data pipelines• Devised and trained 2 CNNs, for binary classification and peak boundary detection ‑ implementedusing bitmasking, using the TensorFlow framework• Implemented Clique detection to align Time‑series data from various parsed files•… Show more • Designed and built ETL data pipelines to process LC‑MS, IMS‑MS and tandem MS/MS data• Used Builder design pattern and SOLID principles to design the ETL data pipeline• Developed base module that can be used in multiple ETL data pipelines• Devised and trained 2 CNNs, for binary classification and peak boundary detection ‑ implementedusing bitmasking, using the TensorFlow framework• Implemented Clique detection to align Time‑series data from various parsed files• Employed Multiprocessing for Parallel computing of massive data streams, thus increasingthe efficiency by 70%. Show less