Led the data-driven decision-making initiatives for Revenue Operations team- Developed the VEO (Visibility-Explainability-Optimization) strategic framework, streamlining RevOps efforts and driving sustainable revenue growth initiatives. - Designed, implemented, and productized a collaborative filtering-based method to identify & locate sales opportunities by leveraging diverse client data. - Formulated Key Performance Indicators (KPIs) critical for the company's profitability goals and implemented a robust monitoring system with anomaly detection alerts, enabling swift corrective actions. Collaborated with cross-functional teams to translate operational gaps into well- defined problems, ensuring that data science projects effectively addressed the organization’s most pressing issues.

Designed a Multimodal Classification Model for e-commerce products - Developed a text classification model with a remarkable 90% accuracy aimed at predicting HS codes from e-commerce product descriptions, skillfully handling the unique challenge of utilizing descriptions tailored for product marketing for the HS code classification. - Engineered a spectrum of models from TFIDF encoding with SVM, to multimodal deep learning architectures (BERT + CNN) that seamlessly combined features from product images and textual descriptions, ultimately elevating performance. Built a volume forecast system for optimizing resource allocation. - Partnered with warehouse operations and network strategy teams in a collaborative endeavor to conceptualize and implement a volume forecast system, reducing the forecast error by over 25% compared to the traditional allocation model. - Engineered a combination of machine learning and time series models, integrating exogenous features, to develop a multi-layered forecasting system. This included a one-week horizon model for labor and transport planning and a one-day-ahead forecast for in-time truck schedule adjustments when necessary.

Led the initiative to adopt computer vision techniques for improving warehouse operations. - Formulated and executed 5+ Proof-of-Concepts showcasing the potential of enhancing operations by leveraging computer vision tools to the stakeholders, resulting in the initiation of multiple projects. - Developed an exception analysis system using YOLO object detection for automated parcel processing, improved categorization and clear accountability leading to a 20% reduction in exceptions during pilot runs. Established mathematical modeling capability within the India team. - Designed and implemented various scenario testing experiments using AnyLogic for the logistics network, providing crucial insights to support business decision-making. - Trained fellow data scientists to develop a modeling mindset and utilize essential tools for modeling and optimizing the logistics network.

Network Optimization - Designed and conducted scenario testing experiments to Analyze the Logistic Network performance during the 2019 peak season. - The project involved extracting data from Snowflake, data pre-processing, Designing the experiment in AnyLogic & deciding the KPIs for the network performance, post-processing the results and summarizing the findings for the stakeholders. Warehouse Operations - POC: Built a computer vision system to detect, reorient & read the barcodes from parcel images.- Deployed a quantized version of model on android device using TensorFlow Lite (TFLite).

Digital Transformation & AI | Mathematical Modeling | Product DevelopmentAdvanced Modeling and AI Integration: – Collaborated in the development of Industrial IoT-based AI solutions to optimize boiler operation and maintenance, contributing to the organization's digital transformation efforts.– Utilized advanced modeling to narrow the design-operational efficiency gap, achieving over 5% efficiency gains and eliminating unplanned shutdowns through predictive maintenance. – Applied mathematical modeling techniques for thermal and structural design of boilers and boiler components, also built simulation models to assess and improve new designs.Product Development and Leadership: – Led cross-functional teams of 10+ members in the development of two innovative product families, managing the end-to-end process from concept design to product launch.– Contributed to the standardization of the entire plant design in conjunction with boiler design, resulting in over 20% reduction in installation costs, while delivering a comprehensive solution. – Implemented design thinking principles to engineer a compact biomass boiler, replacing diesel units in commercial kitchens and achieving a 1,000 metric tons annual CO2 reduction, advancing environmental sustainability and cleaner energy solutions.

Comprehensive Cross-Functional TrainingEngaged in comprehensive cross-functional training across Innovation, Engineering, Proposals, Sales & Service departments to gain a holistic understanding of organizational functions and their collective contribution to shared objectives.