ANURAJ MAURYA

Developed and executed a comprehensive carbon footprint reduction feature for vehicles, encompassing a thorough literature review, establishment of a robust data pipeline, model training, validation, and deployment on an edge device. Utilized advanced libraries such as treelite packages to enhance model efficiency. Proficiently managed feature operations leveraging AWS tools, including Airflow for report generation and EMR for efficient data processing. Developed a robust data pipeline for dataset creation, prioritizing explainability and consistency in predicting Greenzone score. Conducted exploratory analysis and explored ML/DL models such as NN, LSTM, XG Boost, and LightGBM. Opted for Generative Additive Models (GAM) based on KPIs and explainability. Explored diverse metrics for benchmarking to enhance correlation with accident data. Contributed to the Greenzone score API for seamless integration with Cloud. Established an efficient data pipeline utilizing EMR for dataset creation. Conducted an extensive literature survey to develop a CAN-agnostic feature, enabling the computation of the fuel score. Implemented and validated a clustering algorithm to generate the score, ensuring accuracy and relevance in assessing fuel efficiency. Implemented code chunking techniques in the llm to enhance vector search on the Opensearch DB. Introduced automatic validation for SMA answers and optimized SQL generation to expedite the release process. Expanded SMA support to include additional pages, contributing to a more comprehensive and efficient system. Optimized speeding alert module for cost efficiency via AWS Lambda, SQS, and Kinesis stream migration. Implemented a scalable ETL pipeline with AWS EMR for efficient data querying and parsing. Tested, validated, and deployed ML/DL models on resource-constrained edge devices.

Established a company specializing in design and computational analysis for industrial components, as well as the fabrication of machinery and engineering parts.