Vinod Kumar Yerraballi
Associate Data Scientist • Kadapa, Andhra Pradesh • v****************@gmail.com • +91*******335 • linkedin.com/•••••
Professional Summary
Data Science graduate with hands-on internship experience in supervised and unsupervised learning, NLP text classification, and end-to-end project delivery. Proficient in Python-based data engineering, feature engineering, model validation and visualization with proven ability to produce reproducible analyses and stakeholder-ready dashboards. Seeking an entry-level data scientist role to apply statistical foundations and machine learning skills to business problems.
Technical Skills
Programming Languages: Python
Frameworks and Libraries: Pandas,NumPy,scikit-learn,TensorFlow,PyTorch,Matplotlib,Seaborn,Machine Learning,Deep Learning,LLM
Databases: SQL
Data and Analytics: Power Query,Statistical Analysis,Hypothesis Testing,XGBoost,Random Forest,Decision Trees,Regression,Classification,Clustering,Forecasting,Model Validation,Power BI,Streamlit,Web Scraping,BeautifulSoup,Data Cleaning & Transformation,Feature Engineering
Tools and Methodologies: Jupyter Notebook,Git,A,B Testing,Jira,Documentation Standards
NLP & Generative AI: Text Classification,NLP Pipelines,Prompt Experimentation
Work Experience
British Airways
Remote
Data Science Intern
Tech Stack: Python, Pandas, scikit-learn, Streamlit, Jupyter Notebook
- Cleaned and transformed 1,000+ customer feedback records using Python and Pandas, resolving inconsistencies and structuring unstructured text to prepare data for modeling.
- Performed exploratory data analysis and feature engineering on feedback text, extracting signal for sentiment and topic classification and producing candidate features for model training.
- Built and validated a text classification pipeline using scikit-learn (feature extraction, model training, evaluation) and assessed performance with accuracy, precision and recall to ensure alignment with stakeholder requirements.
- Developed 3+ interactive Streamlit dashboards that translated model outputs into visual summaries and class distributions for non-technical stakeholders, enabling faster data-driven decisions.
- Documented preprocessing, feature selection and validation steps in reproducible Jupyter notebooks to ensure handover and repeatable experiments for the analytics team.
- Collaborated with business stakeholders to define annotation guidelines and label schema, improving label consistency and enabling more reliable supervised model training.
The Spark Foundation
Remote
Data Science Intern
Tech Stack: Python, Pandas, scikit-learn, XGBoost, Git, Jupyter Notebook
- Executed 5 supervised and unsupervised ML tasks (regression, classification, clustering) end-to-end, achieving an average model accuracy of ~90% across defined problem statements.
- Conducted structured EDA and hypothesis testing prior to modeling to improve data readiness and reduce iterative rework during validation phases.
- Applied iterative feature engineering and hyperparameter tuning using scikit-learn and XGBoost, improving baseline model performance by up to 40%.
- Validated models using appropriate metrics (R² for regression, confusion matrix and accuracy for classification) and applied cross-validation to ensure generalization.
- Authored 5+ technical write-ups documenting methodology, code and results for non-technical audiences; combined posts reached 500+ views demonstrating clear communication of findings.
- Maintained reproducibility using Git and Jupyter Notebooks and participated in Jira-based workflows to track tasks and deliverables.
Education
Yogi Vemana University, Kadapa
M.Sc. Mathematics • Kadapa, Andhra Pradesh • Feb 2021 – Dec 2022
Coursework: Machine Learning, Statistical Methods, Data Mining, Deep Learning, NLP
Yogi Vemana University, Kadapa
B.Sc. • Kadapa, Andhra Pradesh • Jun 2016 – May 2019
Achievements
- Ranked top 5% in Analytics Vidhya Data Science Contest: Placed in the top 5% among 10,000+ participants by applying structured feature engineering and effective model design.
- Model improvement through experimentation: Improved model accuracy by up to 40% over baseline through systematic experimentation and hyperparameter tuning across projects.
Powered by Drivetube · Create your own profile at drivetube.ai