Suyog Dhage

I’m a current Associate Software Engineer Intern at Coditas in Pune, where I develop backend services with Python and FastAPI. I focus on building clean, maintainable RESTful APIs that follow modular architecture and service-layer patterns.

In my role, I design and implement CRUD operations with proper request validation and structured API responses. I’m also preparing backend systems for future AI/ML integration within scalable microservice-based architectures.

My technical foundation spans machine learning and data science, including supervised/unsupervised learning, deep learning, and end-to-end predictive modeling. I use TensorFlow, Keras, scikit-learn, and OpenCV to turn data and models into practical systems, with an emphasis on model training, evaluation, and performance.

In projects, I built a Deepfake Video Detection System using a CNN with TensorFlow/Keras (85% accuracy) and transfer learning with ResNet50 to cut training time by 60%. I also created a US Visa Approval Prediction System achieving 92% accuracy using scikit-learn, with strong preprocessing, EDA, and hyperparameter tuning via Grid SearchCV.