Skip to content

Jayasimha Obulam

Machine Learning Engineer • Kadapa, Andhra Pradesh • o**************@gmail.com • +91*******539 • linkedin.com/••••• • drivetube.ai/•••••

Career Objective

Machine Learning Engineer with 0 years of experience developing end-to-end ML and deep learning solutions using Python, TensorFlow and XGBoost; built blood cell classification (95.8% accuracy) and credit-card fraud detection models (ROC-AUC 0.972). Seeking entry-level ML/AI roles to apply model-building, data preprocessing and evaluation skills.

Education

Kandula Obul Reddy Memorial College of Engineering
B.Tech in Computer Science Engineering (Artificial Intelligence and Machine Learning specialization) • Kadapa, Andhra Pradesh • 2022 – 2026

Technical Skills

Programming Languages: Python
Frameworks and Libraries: TensorFlow,NumPy,Pandas,scikit-learn
Databases: SQL
Data and Analytics: Supervised Learning,Feature Engineering,Class Imbalance Handling,XGBoost,Model Selection,Data Preprocessing,Data Cleaning,Exploratory Data Analysis
Tools and Methodologies: Jupyter Notebook,Visual Studio Code,Git,GitHub
Deep Learning: Convolutional Neural Networks,Transfer Learning

Projects

Hematovison — Advanced Blood Cell Classification Using Transfer Learning
  • Developed a CNN-based blood cell classification model using transfer learning and TensorFlow to improve diagnostic image classification.
  • Performed image preprocessing and extensive data augmentation to increase effective dataset size and reduce overfitting.
  • Evaluated model with accuracy and F1-score and used confusion matrix analysis to iterate on class-specific improvements.
  • Achieved 95.8% classification accuracy on validation data and produced reproducible notebooks and model evaluation artifacts.
Credit Card Fraud Detection Using State-of-the-Art Machine Learning
  • Built machine learning pipelines to detect fraudulent credit-card transactions using feature engineering and imbalance-handling techniques.
  • Trained and compared models including XGBoost and scikit-learn classifiers; selected models based on ROC-AUC and recall for fraud detection.
  • Evaluated model performance yielding precision 0.616, recall 0.867 and ROC-AUC 0.972; reported model diagnostics and business trade-offs.
  • Reported a final classification outcome and produced evaluation reports indicating high detection recall; documented code on GitHub.

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