Ranjith Kumar G

India

@ranjithkumar28

Data Scientist

Badges

Problem Solving
CPP
Java
Python
Days of Code
Days ofStatistics
Sql

Certifications

Work Experience

  • Data Scientist

    Saviynt•  August 2024 - Present•  Bangalore

    • Developing a description generation application for entitlement values using metadata and historical data. Leveraging Natural Language Processing, Unsupervised Learning, and Generative AI LLMs from AWS Bedrock to effectively create meaningful descriptions.

  • Systems Enginner

    Tata Consultancy Services Ltd•  May 2021 - August 2024•  Bangalore

    • Developed and deployed a Python-based ML application using Flask, Pandas, Scikit-Learn, and Azure to forecast Azure resource utilization. The app analyzes forecasted data to identify frequent idle and underutilized time slots. • Implemented a Python-based ML regression model using Pandas and Scikit-Learn to predict future financial operation costs for Azure subscriptions, resource groups, and resources. Implemented a classification model to categorize usage patterns in forecasted costs as normal, deficient, or excessive. Deployed the model as a REST API on Azure. • Developed a RAG-based GenAI Manufacturing Copilot using FastAPI, Azure OpenAI, Azure AI Search, Azure Document Intelligence, and LangChain. It incorporates multi-document parsing, chunking, embedding creation, vector storage, effective QnA, follow-up question generation, citation generation, and relevant images retrieval. • Developed a Python package for repetitive GenAI tasks like parsing, chunking, embedding creation, embeddings storage, QnA, citation generation, relevant images retrieval, and follow-up questions generation. Deployed on Azure Artifacts for secure access.

  • Artificial Intelligence Intern

    Object Automation Software Solutions Pvt ltd•  June 2019 - July 2019•  Chennai

    • Created highly accurate tree-based supervised machine learning model for weather forecasting. • Worked on hyper parameter optimization of the weather forecasting model to acheive better accuracy. • Created CNN model for multi-class image classification for MNIST dataset .

Education

  • Hindustan university

    Computer Science & Engineering, B.Tech•  2017 - 2021

Skills

Python(Advanced)
Machine Learning
SQL
Generative AI
Deep Learning
Natural Language Processing