Shekhar Samudre

About Shekhar Samudre

As an experienced AI and machine learning specialist, I am developing a Retrieval-Augmented Generation (RAG) solution for medical and policy documents. Leveraging OpenAI\'s advanced language models, I implemented recursive character splitting and semantic chunking to enhance document processing and information retrieval. My solution incorporates FAISS vector search for efficient and accurate vector storage and retrieval, ensuring high performance and scalability. Additionally, I utilized the Langchain framework to streamline the integration and deployment of the AI models.I employed state-of-the-art models, such as Roberta and Alibaba\'s base models, for semantic chunking, achieving superior accuracy and contextual understanding. This sophisticated approach enables precise extraction and summarization of information from complex documents, driving significant improvements in data accessibility and utility. To further augment this solution, I utilized AWS Textract for data extraction from scanned PDFs, ensuring that even non-digital documents are seamlessly integrated into the workflow.My recent experience with UNICEF as a researcher and analytics consultant involved analyzing public domain and confidential data. I used tools such as MS-Excel, Power BI, QGIS, and GeoDa for mapping and applied machine learning algorithms like linear and logistic regression. I also analyzed the childbirth seasonality of Gujarat using the ARIMA model for time series analysis. I have consulted with the Ministry of Women & Child Development (GOI), participated in the policy think tank RJMCHN for the Government of Maharashtra, and managed the Women and Child Development Department of Nashik division under the Divisional Commissioner of Nashik.I am well-trained in supervised, unsupervised, ensemble, and hyperparameter tuning. Additionally, I have expertise in AWS Cloud and SageMaker and have completed course projects on ANN, CNN, RNN, and LSTM.My expertise spans the entire AI development lifecycle, from model selection and data preprocessing to deployment and optimization, ensuring robust and reliable solutions tailored to specific business needs.In my spare time, I like to play Cricket & Volleyball, watch Movies, surf Netflix & Travelling. I love to interact with people & understand their perspectives on different things. I have a keen interest in such roles that would give a mixed flavour of all my learnings.

Experience

1. Lead Analyst

   ANZ

   Mar 2025 — Present · Bengaluru, IN

   Credit Risk Modelling - Stress Testing Models

Education

• Indian Institute of Technology, Bombay

  Master of Technology - MTech, Technology and Development

  2019 — 2021

• National Institute of Technology, Tiruchirappalli

  Master of Technology - MTech, Non destructive testing

  2018 — 2019

• Pravara rural college of engineering

  Bachelor of Engineering - BE, Mechanical Engineering

  2012 — 2016