Harshvardhan Patil

+1-919-***-**** ad2dfb@r.postjobfree.com ad2dfb@r.postjobfree.com www.linkedin.com/in/harshvardhan-patil14 EDUCATION

Master of Computer Science August 2023 - April 2025 North Carolina State University North Carolina, USA Bachelor of Technology, Computer Science and Engineering July 2018 - July 2022 D. Y. Patil College of Engineering and Technology Maharashtra, India GPA: 8.88/10 SKILLS

● Programming Languages: C, C++, Python, Dart, JavaScript

● Frameworks: Flutter, Flask, Tensorflow, PyTorch, Keras

● Programming Libraries: Numpy, Pandas, SciKit, Data visualization libraries, OpenCV

● Databases: MYSQL, SQL Server, Firebase

● Visualization Tools: PowerBI

● Other Tools: MS-Excel, Google Sheets

WORK EXPERIENCE

Tata Consultancy Services Systems Engineer February 2023 - July 2023

● Worked with a team to design and develop an SQL database. Resolved multiple problems that appeared in the SQL stored procedures and functions through debugging. Optimized complex SQL queries, reducing loading times by 10% and enhancing dashboard performance.

● Performed data analysis by analyzing big-data collected from IoT sensors, performed data visualization tasks utilizing Tableau, resulting in comprehensive insights, pattern detection, and anomaly identification, enhancing decision-making processes and improving operational efficiency by 25%.

Softmusk Info Pvt. Ltd. Data Science Intern January 2022 - May 2022

• Built a highly accurate Convolutional Neural Network (CNN) model with SHA-1 algorithm to authenticate multimedia content

(images, videos, audio, PDFs, and text files); achieved an exceptional precision rate of 98.25%

• Developed and deployed a robust face intrusion detection model using OpenCV and Convolutional Neural Network in Python, achieving 92% accuracy and reducing false alarms by 23% within a real-time surveillance system.

• Implemented a Convolutional Neural Network (CNN) based plant disease detection model in Python, achieving 89% accuracy in identifying diseased plant images and providing actionable recommendations for remedial actions based on an extensive training dataset.

PROJECTS

Wolf Parking Management System Development

• Database Architecture & Implementation: Led BCNF and 3NF optimized design with SQL triggers/functions for seamless operations like adding/updating/deleting driver and parking info, ensuring data integrity.

• Engineered and implemented atomicity and consistency protocols in SQL transactions for permits, vehicle data, and citation processes, resulting in a 40% reduction in data discrepancies and enhancing operational efficiency.

• Advanced Reporting Integration: Employed SQL queries with efficient indexing for detailed insights into citations, zone listings, violations count, and space availability in specific parking lots.

• Technologies used: SQL, Java (JDBC)

Smart Attendance system using Advanced Face Recognition

● Engineered a effective machine learning model leveraging deep learning algorithms to accurately detect and recognize faces, enabling automated attendance tracking for 50+ students, reducing manual processing time by 90% and eliminating errors.

● A system which uses centralized camera placed in the classroom, to capture images and then detects the faces. Through face recognition and CNN techniques their attendance was marked.

● The model is trained on three images per student. Thus, the newly captured photos are recognized by the system and attendance is marked in the database.

● Technologies used: OpenCV, Face recognition libraries in python. Agricultural cropping patterns: Prediction of crops pattern across Maharashtra using machine learning.

● Devised a data-driven predictive model leveraging the Random Forests algorithm to optimize crop patterns; identified key factors influencing crop selection, leading to a 27% increase in agricultural productivity and a 15% boost in profitability.

● Diverse information, including meteorological conditions, soil properties, water availability, and market demand, were curated and merged to ensure compatibility for input into the Random Forests model.

● Cross-validation and hyperparameter tuning were used to fine-tune the Random Forests model, resulting in a precision and recall score for accurate crop prediction.

● Technology Used: Python

AWARDS AND HONOURS

● Awarded as Best Outgoing Member of Coding Club.

Contact this candidate