Skip to content

HARI NARAYANA RAO CHEPURI

Aspiring Data Scientist / ML Engineer • Hyderabad, IN • h***************@gmail.com • +91*******063 • linkedin.com/••••• • github.com/••••• • github/•••••

Professional Summary

Computer Science graduate and data science trainee with hands-on experience building end-to-end ML, DL and NLP solutions. Proficient in Python, SQL, feature engineering, model development and deployment using FastAPI, Flask, Docker and MLflow. Experienced with Generative AI, RAG, LangChain and cloud platforms (Azure, AWS, GCP). Seeking an entry-level Data Scientist / ML Engineer role to apply data-driven approaches and production-aware MLOps practices to solve business problems.

Technical Skills

Programming Languages: Python
Web Technologies: REST APIs
Frameworks and Libraries: Pandas,NumPy,Matplotlib,Seaborn,TensorFlow,Keras,PyTorch,FastAPI,Flask,LangChain
Databases: SQL
Cloud and DevOps: Docker,MLflow,Model monitoring,Azure,AWS,Google Cloud Platform,Data pipelines
Data and Analytics: Power BI,Regression,Classification,Ensemble Methods,XGBoost,SMOTE
Tools and Methodologies: Prompt engineering,RAG Retrieval-Augmented Generation,LLM APIs,Vector databases,Agentic workflows,Git
Deep Learning & NLP: LSTM,NLP feature engineering

Work Experience

Vcube Software Solutions
Hyderabad, IN
Data Science Trainee
Feb 2025 – Jul 2025
Tech Stack: Python, Pandas, NumPy, scikit-learn, XGBoost, TensorFlow, PyTorch, Flask, FastAPI, Docker, MLflow, LangChain, Vector databases, Git
  • Performed exploratory data analysis and statistical analysis on real-world datasets using Python, Pandas and NumPy to identify key signal and clean data for modeling.
  • Engineered features and built supervised models (regression, classification, ensemble methods) with scikit-learn and XGBoost, applying hyperparameter tuning and cross-validation to improve generalization.
  • Developed deep learning and sequence models (LSTM) and experimented with TensorFlow/Keras and PyTorch for time-series and NLP tasks to prototype production-ready solutions.
  • Designed and implemented REST API endpoints using Flask and FastAPI, containerized applications with Docker and tracked experiments and models using MLflow for reproducible deployments.
  • Applied Generative AI techniques, prompt engineering, LangChain and RAG with vector databases to build retrieval-augmented and agentic prototypes for industry problem statements.
  • Adopted MLOps practices including model monitoring, automated data pipelines and Git-based workflows to ensure reproducible training, deployment and handover to engineering teams.

Projects

Stock Price Prediction | Dec 2025 – Apr 2026
Tools Used: Python, LSTM, Yahoo Finance API, MLflow, Matplotlib, Pandas, NumPy
  • Built an end-to-end time-series forecasting pipeline using LSTM to predict next-day closing prices, ingesting real-time data from Yahoo Finance API.
  • Performed data preprocessing and normalization (MinMaxScaler), implemented a 60-day sliding window approach and tuned hyperparameters to enhance forecast stability.
  • Packaged training and inference steps into modular pipeline components, logged experiments and models with MLflow, and visualized actual vs predicted trends using Matplotlib.
Autism Prediction using ML | Aug 2025 – Nov 2025
Tools Used: Python, XGBoost, SMOTE, Scikit-learn, Matplotlib, Pandas
  • Designed a predictive model to identify autism risk from clinical and demographic features, performing preprocessing and class balancing with SMOTE.
  • Trained and optimized tree-based models (Decision Tree, Random Forest, XGBoost) with hyperparameter tuning and validated results to prepare for deployment.
Customer Churn Prediction | Feb 2025 – Jul 2025
Tools Used: Scikit-learn, SQL, Power BI, Pandas
  • Built a churn prediction model on 200K+ customer records using feature engineering and model validation to identify at-risk customers, achieving 85% accuracy on validation data.
  • Developed automated Power BI dashboards to monitor churn KPIs, reducing manual reporting time by ~70% and enabling data-driven retention strategies.
Fake News Detection System | Dec 2023 – May 2024
Tools Used: NLP, Scikit-learn, Docker, Flask, Pandas
  • Built an NLP-based multi-class fake news classifier by processing and feature-engineering 50K+ news articles, achieving 89% precision using ensemble techniques.
  • Containerized the model and web interface with Docker and deployed via Flask, and produced feature-importance analysis to improve interpretability for end users.

Education

Vel Tech University
B.Tech in Computer Science Engineering • Chennai, IN • Jun 2020 – May 2024
Sri Gayatri Junior College
Secondary Education (MPC) • Tirupathi, IN • Aug 2018 – Mar 2020

Certifications

AI – Data Scientist — Reliance Foundation Skilling Academy • Jul 2026
AI – Machine Learning Engineer — Reliance Foundation Skilling Academy • Jun 2026
Data Science for Everyone — Reliance Foundation Skilling Academy • Jun 2026
Python for Data Science — Saylor Academy • Mar 2026
Python Programming — Microsoft / Skill India Digital Hub • Mar 2026
Data Analytics Job Simulation — Deloitte Australia / Forage • Feb 2026

Powered by Drivetube · Create your own profile at drivetube.ai