GAURAV SANWAL
Machine Learning Enthusiast | Python Developer
Mohali, Punjab, India, 140413 | gauravsanwal88@gmail.com | +91 8859696623
LinkedIn: https://www.linkedin.com/in/gaurav-sanwal-3263b3221 | GitHub: https://github.com/gaurav022003/gaurav022003
Professional Summary
Driven to push the boundaries of artificial intelligence and machine learning to tackle complex real-world challenges. Currently
pursuing a Bachelor's degree in Computer Science & Engineering with a specialization in Artificial Intelligence and Machine
Learning at Chandigarh University.
EDUCATION
Chandigarh University September 2021 - June 2025
Bachelor's, Computer Science and Engineering AIML (Hons.)
SKILLS
• Programming Languages: Python, C++, Java
• Machine Learning: TensorFlow, Keras, Scikit-learn, PyTorch
• Deep Learning: Neural Networks, Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN)
• Computer Vision: OpenCV, YOLO
• Development Tools: Git, Docker, Jupyter, Streamlit, FastAPI
CERTIFICATIONS
• Advanced Machine Learning on Google (Specialization) Verified Certificate
• Introduction to Computer Vision and Image Processing (IBM) Verified Certificate
PROFESSIONAL EXPERIENCE
AI Engineer, WictroniX
Jun 2023 - Aug 2023
• Developed a custom algorithm to estimate vehicle speed using drone footage with 80% accuracy and an inference speed
of 10 FPS using a custom-trained YOLO V8 model and ByteTracker.
• Enchanced road traffic management and road audit reporting.
• Collaborated on an AI project with the University of Munich, Road Ministry Surat, and NIT Surat.
PROJECTS & OUTSIDE EXPERIENCE
AI-Supported AICTE Approval Process Portal (SIH-2024 Finalist)
Tools: Fastapi, Docker, Git, Opencv
• Engineered OCR-based document verification workflows, automating data extraction and signature validation processes.
• Designed and implemented AI-powered data validation modules, enhancing efficiency and accuracy in document
processing.
• Collaborated with cross-functional teams to integrate chatbots for automated guideline assistance and FAQ management.
• Streamlined approval workflows, including scrutiny checks, expert reviews, and compliance with AICTE regulatory
standards.
Hierarchical Multi-Class Classification
Tools: TensorFlow, Streamlit
• Designed a new architecture to leverage the hierarchy between different classes, achieving a 5% increase in accuracy
compared to other state-of-the-art architectures.
• Published a research paper in an IEEE Scopus-indexed conference at IIT UNA, achieving acceptance in a highly
competitive environment with a 26% acceptance rate.
Plant Disease Detection
Tools: TensorFlow, FastAPI, CNN
• Utilized the Plant Village dataset with 15 classes of diseases and over 20,000 images.
• Achieved 85% accuracy and implemented an API using FastAPI, reducing inference time to approximately 30 seconds,
enhancing responsiveness and efficiency in detection.
Automated Number Plate Recognition System
Tools: OpenCV, Neural Networks
• Trained a model using YOLO V5 with a dataset from Roboflow.
• Achieved 98.4% mAP, 98.2% precision, and 96.7% recall with an inference speed of 15 FPS.
• Improved recognition speed by 20%.
Achievements
• AI Mentor in ThinkTank Ideathon by AIML Society, Chandigarh University, guiding 20 teams in the final round.
• Ranked in the top 30 out of 700 teams in the DevBhoomi CyberSecurity Hackathon at IIT Roorkee.
• Final Round Participant in the hackathon at UPES Dehradun, securing a spot among the top 20 out of 500 teams.
• Achieved a top 10 position in the Hackoverflow hackathon at Chandigarh University .
