About

I am Jothi Lakshmi H, a recent graduate with a B.Tech in Electronics and Communication Engineering from Thamirabharani Engineering College, Tirunelveli. During my studies, I completed several projects in AI and ML, including real-time object detection using YOLOv8 and a multiclass fish image classification system using PyTorch. I also gained hands-on experience through internships and coursework in machine learning, deep learning, and computer vision. My strongest skills include Python, SQL, Streamlit, YOLOv8, OpenCV, PyTorch, TensorFlow, and Power BI. I am now looking for my first full-time role as an ML Engineer.

Education

1. Thamirabharani Engineering College, Tirunelveli

   B.E. - Electronics and Communication Engineering

   2020 – 2024

   Relevant Coursework: Machine Learning, Deep Learning, Computer Vision

2. IIT-M GUVI Pravartak

   IITM-M Artificial Intelligence and Machine Learning Program

   2024 – 2025

Projects

• Military-Soldier-Safety-and-Weapon-Detection-using-YOLO-and-Computer-Vision

  • Python
  • YOLOv8
  • OpenCV
  • Deep Learning
  • Streamlit

  Designed and developed a real-time multi-class object detection system to identify soldiers, weapons, vehicles, civilians, and trenches for battlefield threat awareness. Fine-tuned YOLOv8m using transfer learning on a custom-annotated dataset, improving convergence speed and detection accuracy.

• Multiclass-Fish-Image-Classification-deep-learning

  • Python
  • PyTorch
  • CNNs
  • Transfer Learning
  • Streamlit

  Built a multiclass image classification pipeline to identify 11 fish species using CNNs and pretrained architectures. Applied transfer learning using a pretrained MobileNetV2 as a feature extractor, achieving 98.7% accuracy, 96.5% precision, 97.5% recall, and 97% F1-score on the test set.

• PowerPulse-Household-Energy-Usage-Forecast

  • Scikit-learn
  • Machine Learning
  • Feature Engineering
  • Data Visualization

  Developed predictive models including Linear Regression, Random Forest, and Gradient Boosting to forecast household electricity consumption. Implemented feature engineering, outlier handling, and data preprocessing to enhance forecasting accuracy and model stability.