Skip to content

Jeevan Challa

B.Tech Student (AI & ML) • Hyderabad, Telangana, India • j************@gmail.com • +91*******626 • linkedin.com/••••• • drivetube.ai/•••••

Career Objective

Machine Learning Engineer with 0 years of experience building end-to-end ML and deep learning solutions, focused on model development, evaluation, and deployment using Python, Scikit-learn, TensorFlow, PyTorch, Docker, FastAPI, and AWS.

Education

Nalla Malla Reddy Engineering College
B.Tech in CSE (AI & ML) • Hyderabad, India • 2023 – 2027
Sri Chaitanya Jr. College
Intermediate (Maths, Physics and Chemistry) • Hyderabad, India • 2021 – 2023
Balaji Techno School
Secondary Education (CBSE) • Warangal, Telangana, India • 2020 – 2021

Technical Skills

Programming Languages: Python
Frameworks and Libraries: Scikit-learn,TensorFlow,PyTorch,OpenCV,Pandas,NumPy,Matplotlib,Seaborn,FastAPI,Flask
Databases: SQL,MySQL,MongoDB
Cloud and DevOps: Docker,Streamlit,AWS EC2,AWS S3
Data and Analytics: Regression,Classification,Clustering,Feature Engineering,XGBoost
Tools and Methodologies: Git,GitHub
Computer Vision & NLP: Transfer Learning,CNNs,NLP

Projects

Hospital Patient Readmission Prediction
Tools Used: Python, Scikit-learn, XGBoost, Pandas, Feature Engineering, Streamlit
  • Ingested and cleaned an 8,000-record patient dataset, handling missing values, categorical encoding, and class imbalance to prepare features for model training.
  • Engineered clinical and utilization features and applied cross-validation to select a robust model pipeline using XGBoost and Scikit-learn.
  • Optimized model recall through threshold tuning and feature selection, achieving 89.73% recall on the validation set for high-risk readmission detection.
  • Implemented model evaluation using precision-recall and ROC analysis to validate clinical utility and reduce false negatives for readmission alerts.
  • Packaged the model into a reproducible inference pipeline and exposed predictions via a Streamlit interface for demonstration to stakeholders.
SAR-Optical Cross-Modal Transformer for Camouflage Detection
Tools Used: Python, PyTorch, Transformers, Cross-attention, Flask, Docker
  • Designed a cross-modal transformer combining SAR and optical inputs with cross-attention to improve detection of camouflaged objects in remote-sensing imagery.
  • Implemented model architecture in PyTorch, leveraged transfer learning on convolutional backbones and custom transformer blocks for cross-modal fusion.
  • Trained and validated on multimodal datasets, achieving 94%+ accuracy through augmentation, balanced sampling, and careful hyperparameter tuning.
  • Built an inference API using Flask and containerized the application with Docker for consistent deployment and evaluation across environments.
  • Analyzed failure cases and refined attention weighting to reduce false positives in cluttered environments, improving precision for operational scenarios.

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