EXPERIENCE
RUTGERS NEW JERSEY MEDICAL SCHOOL, FREUNDLICH LAB (New Brunswick, NJ, USA)
Research Assistant (Data Science)
In my role as a Data Scientist specializing in cheminformatics, I significantly improved the efficiency and accuracy of our analytical processes related to organic molecular data. My first major achievement was reducing the analytical runtime by 40%. This was accomplished through meticulous molecule standardization, data deduplication, and extensive data preprocessing and cleansing, employing standard cheminformatics functions. These optimizations streamlined the analysis, allowing for quicker and more reliable insights into molecular data sets, facilitating faster decision-making in research projects.
Building on this, I developed a machine learning (ML) pipeline that utilized non-linear regression techniques, enhanced with optimized ridge and lasso regularization methods. This advanced pipeline was designed specifically to address the complexities of the drug discovery process. Through this initiative, I achieved a 15% improvement in prediction accuracy and a significant 20% reduction in false positive rates. These enhancements were crucial in accelerating the identification and validation of potential drug candidates, thus streamlining the drug development lifecycle and contributing to more efficient and effective drug discovery efforts.
LYONS INFORMATIONS SYSTEMS (Edison, NJ, USA)
Data Analyst Intern
In my capacity as a Data Scientist, I spearheaded an analytical initiative leveraging over 4 million data points to conduct advanced statistical analysis, pinpointing the factors responsible for discrepancies between expected and actual shipment times. This rigorous analysis provided critical insights, leading to strategies that projected an 8% enhancement in supply chain efficiency.
Further elevating product satisfaction, I employed a custom, hyperparameter-tuned BERT (Bidirectional Encoder Representations from Transformers) algorithm to analyze consumer feedback sentiment. This deep learning approach was instrumental in decoding the underlying reasons for product returns, culminating in a strategic foresight that anticipated a 14% decline in return rates. The success of this analysis underscores the potent application of NLP techniques in extracting actionable business insights from large-scale consumer data.
Complementing these analytical feats, I designed and implemented custom dashboards in PowerBI, adeptly integrating complex datasets to elucidate customer loyalty metrics and identify pivotal areas for supply chain and product quality enhancements. This initiative was projected to lead to significant cost reductions, with anticipated savings of around 30%. These accomplishments reflect my comprehensive approach to data science, blending sophisticated statistical methods, cutting-edge machine learning algorithms, and robust data visualization tools to drive substantial business improvements.
CAPGEMINI (Pune, MH, India)
Senior Software Engineer, Analyst
As part of a dynamic team, I contributed to the development of a budget management application, focusing on both backend and frontend functionalities. Utilizing Python and SQL, I helped construct a robust backend system, while integrating with React on the frontend to create a responsive and user-friendly interface. Our efforts led to a notable 20% reduction in latency, significantly enhancing data processing efficiency and improving the overall user experience.
In the realm of data analysis, I curated sophisticated algorithms, including non-linear regression and multi-class classification, to meticulously analyze user data. This analysis enabled the generation of personalized budget model options, effectively utilizing novel user inputs along with legacy spending patterns. This approach resulted in a substantial improvement, boosting savings per user by 37%.
To further refine the system's efficacy, I engaged in predictive modeling and rigorous A/B testing, focusing on the integration of recurrent costs. This optimization effort successfully enhanced the accuracy of our financial predictions by 7%. Additionally, I engineered a neural network model designed to forecast future spending behaviors for individual users. This model dynamically updates their budget plans, ensuring continuous alignment with their financial goals and spending habits. Through these initiatives, I played a pivotal role in creating a data-driven application that empowers users with real-time, personalized financial insights, leading to better budget management and financial planning.
SHREE POLYMERS (Ankleshwar, GJ, India)
Data Analyst Intern
In my role, I focused on optimizing the Bill of Materials (BOM) and inventory management processes through advanced statistical analysis, achieving a significant reduction in operational delays and material shortages. By meticulously analyzing historical data and identifying patterns, I streamlined the BOM and inventory systems, resulting in a 30% decrease in delays and a 15% reduction in material shortages. This not only improved production efficiency but also enhanced the overall supply chain reliability.
Additionally, I leveraged statistical inference techniques on Quality Assurance (QA) feedback to gain deeper insights into product quality issues. This analysis enabled the identification of key factors contributing to product returns. By addressing these quality concerns, I facilitated a 25% reduction in return rates, thereby increasing customer satisfaction and trust in our products. This work involved close collaboration with the QA and product development teams to implement targeted improvements that directly addressed the identified quality gaps, showcasing the importance of data-driven decision-making in enhancing product quality and operational efficiency.
