Prasad G.

Key Skills

Freelancer Overview

Prasad Dattatray Ghadge +91 8484077268 linkedin.com/in/prasad-ghadge-499564260 [email protected] github.com/prasad999999 Pune, Maharashtra - 411048 EDUCATION Vishwakarma Institute of Information Technology, Pune 2022 - 2026 B.Tech, Computer Science and Engineering (AI) - CGPA - 8.32 Pune, Maharashtra Dr VG Alias Kakasaheb Paranjape Junior College, Rahimatpur 2020 - 2022 Higher Secondary Education - Percentage - 80.83% Rahimatpur, Maharashtra Universal Knowledge School, Satara 2020 Secondary Education - Percentage - 91.83% Satara, Maharashtra WORK EXPERIENCE AlgoAnalytics 06/2025 – 12/2025 Full Stack Intern Pune, India • Built an AI anomaly detection system using PatchCore & WideResNet, achieving 97% accuracy and reducing manual defect inspection by 80%. • Automated customer schedule analysis with GPT-based Excel processing and interactive Plotly dashboards. Eduplus Campus 06/2024 – 12/2024 AI Software Developer Intern Pune, India • Developed a real-time classroom attention monitoring system integrating EEG signals from the MindLink headband with computer vision–based facial recognition for automated attendance and engagement tracking. • Developed an ensemble of CNN models trained on 35,000+ FER images, achieving 74% accuracy and generating downloadable attention reports for educators. PROJECTS AI-Powered Finance Platform | Next.js, React, Supabase, Gemini API, Shadcn UI • Built a Next.js + Supabase full-stack finance platform with budget tracking,

Labeling Experience

Crop Health and Weed Detection using Drones

ImageBounding Box

Annotated aerial agricultural images captured by drones for training object detection models as well as plant disease detection models. The project focused on identifying crops, weeds, and other field anomalies and plant diseases that can be identified from plat leaves for crop health monitoring. Responsibilities included drawing bounding boxes around target objects, reviewing annotations for accuracy, and maintaining consistent labeling standards across the dataset. The labeled data was used to train and evaluate YOLO-based computer vision models for automated weed detection and for plant disease detection we used pretrained Resnet50 model through transfer learning.

2024 - 2025