Parvathy M R

About Parvathy M R

As a results-driven Data Scientist with a strong academic foundation in Machine Learning and Computing, I bring hands-on experience in developing and deploying advanced AI and ML solutions within industry and research environments. My expertise spans Generative AI, Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), NLP, Time-Series Forecasting, and Physics-Informed Machine Learning.Currently at Robert Bosch India, I focus on Generative AI adoption and intelligent AI agents. My notable works include: Developing a retrieval-augmented generation (RAG) system that enhanced internal document intelligence, improving search accuracy by 25% and cutting response time in product teams.Building an LLM-powered exploratory data analysis assistant, enabling natural language–to–automatic code generation and execution, boosting analyst productivity by 40%, and democratizing access to advanced analytics. Additionally, designed an AI-driven career mapping system powered by semantic embeddings and similarity search, accelerating workforce planning and improving internal mobilityBeyond GenAI, I have delivered impactful solutions in time-series forecasting (LSTM–XGBoost ensemble), physics-informed ML models, and e-commerce data mining, achieving measurable improvements in forecasting accuracy, resource optimization, and ROI.I hold a Master of Technology in Machine Learning and Computing from the Indian Institute of Space Science and Technology. With a blend of strong academic grounding and hands-on experience, I approach complex challenges with analytical rigor and innovative thinking. Driven by a passion for continuous learning and emerging technologies, I strive to contribute to high-impact projects and collaborate with visionary teams to advance cutting-edge solutions.

Experience

1. Assistant Manager

   Bosch India

   Jul 2023 — Present · Bengaluru, IN

   Virtual Modeling of Sensors- Spearheaded the development of a virtual temperature sensor model for 2-wheeler engine blocks, leveraging data-driven and physics-based approaches to predict engine temperature within ±5ºC, aiming to replace costly physical sensors- Explored and optimized multiple modeling approaches including map-based models (2D and 3D calibrated maps), thermodynamics-based thermal models, and machine learning models- Integrated hybrid modeling strategies combining thermodynamics-based thermal models with deep learning architectures (LSTM, TCN) to capture complex temporal patterns and improve prediction accuracy- Validated models on real-world dynamic driving datasets, addressing challenges like cumulative error and proposing innovative error correction mechanisms for enhanced reliability.

Education

• GOVT. RAJIV GANDHI INSTITUTE OF TECHNOLOGY KOTTAYAM

  B.TECH

  2014 — 2018

• Indian Institute of Space Science and Technology

  Master of Technology - MTech

  2021 — 2023