R SRUTHI PARVATHA

Data Scientist with 2+ years of hands-on experience in developing machine learning models, performing data analysis, and building end-to-end ML solutions. Proficient in Python, Pandas, Scikitlearn, machine learning models. Skilled in implementing machine learning and deep learning models like linear regression, kNN, decision trees, random forests, XGBM, LightGBM, NN, CNN etc. Experienced in AWS services like S3, Sagemaker AI, Bedrock, Lambda, API Gateway and deploying ML models in AWS. Built interactive ML applications with Streamlit and Flask frameworks. Successfully developed multiple projects in areas like content & image generation using Bedrock, healthcare insurance analysis, demand forecasting, price prediction, customer segmentation. Proactive and enthusiastic in expanding AI/ML skillset with a focus on Gen AI, Large Language Models (LLMs). Developing practical experience in fine-tuning LLMs using LangChain, HuggingFace etc, seeking to apply these skills in a professional setting to contribute to innovative language-based solutions. Passionate about applying data-driven insights and eager to secure a job in the domain of data science where I can leverage my analytical and technical expertise to drive impactful solutions.

RAG Application with LangChain and Open AI for documents, Developed a Retrieval Augmented Generation (RAG) Application on LLM using LangChain framework for retrieving relevant data from documents such as developer documentation (AWS), books etc, storing it in vector store (ChromaDB) and providing it as a prompt to Open AI LLM to generate customised response to user input. Content Generation on a user given topic using AWS Bedrock, In this project, user can submit a topic of choice which requires content generation as an API request via Postman application. Upon receiving API request, AWS API Gateway will route the request to AWS Lambda which invokes AWS Bedrock's LLM namely Llama3 for text content generation and this content is provided as the API response to user. Image Generation using AWS Bedrock, User can submit a detailed description of image as text with an API request via Postman application. Upon receiving API request, AWS API Gateway will route the request to AWS Lambda which invokes AWS Bedrock's LLM namely Titan for image generation and the generated image is stored in S3 bucket for user download/access. Healthcare Insurance Analysis Project, Analysed healthcare insurance dataset to provide insights for cost management, KPI (Cost Variation by Medical Condition, Hospital and City Tier Analysis etc). Developed a predictive model for healthcare insurance costs, achieving 15% improvement in prediction accuracy. Key activities involved - data preprocessing, data visualization, exploratory data analysis, hypothesis testing, feature engineering & selection, model training and model testing in jupyter notebook. Apollo Project, Worked in the discipline of APIs & Integration, as an Integration Developer. Designed & developed several Boomi interfaces to synchronise data in different formats like JSON, CSV, XML, TSV etc from multiple boundary systems to Oracle ERP cloud, performed end to end testing of these integrations and had a successful go-live of the project. Time-Series Forecasting with Amazon SageMaker Autopilot, Analysed food demand dataset uploaded in AWS S3, created and trained an optimized model using SageMaker autopilot, deployed trained model as a real-time endpoint using SageMaker endpoint, invoked and tested model using test data hosted in S3. Signature Based Radar Target Classification (DRDO), The aim of this project is to perform a comparative study of AI based techniques for Radar target classification (Fighter aircrafts, helicopters, UAVs) using HRRP data. Other Machine Learning & Data Science Projects, Customer Segmentation using KMeans Clustering, Product Recommendation System, Health Diagnosis Model for 3 diseases using ML models, NFL Football Statistics Application using Streamlit, Sales Analysis using Power BI, Olympic Winner Analysis using Power BI. TrendyTech JobPortal, Implemented various new functionalities for both admin and user end of TrendyTech JobPortal application built using MERN Stack. Off-Premises Asset Management using IoT Edge, Based on the concept of Predictive Maintenance in Smart Industries. The real-time health of industrial assets is monitored using sensors, Raspberry Pi and Machine Learning algorithms.

09/05/25, Bengaluru, R Sruthi Parvatha

• Applause Award, 2024, Deloitte
• Outstanding Performance Award, 2022, Deloitte
• First Place in National Level Competitions, 2020
• Prerna 2020
• Vertechx 2020
• Top 35 finalists, 2020, Mitsubishi Electric cup
• Selected for Semi Finals, 2019, IICDC 2019, Transformer Health Monitoring System

• PGP Data Science, Purdue University (Simplilearn)
• AWS Certified Cloud Practitioner 2023
• Machine Learning, Data Science and Deep Learning with Python
• Machine Learning with Python, A Practical Introduction - IBM(edx)
• Neural Networks and Deep Learning(Coursera)
• Tutorials on TensorFlow 2.0 from freeCodeCamp.org
• Dell Boomi Professional Developer 2022
• Dell Boomi Professional API Design 2022