Pratham Saurabh
Cybersecurity Engineer & AI Developer
Third-year Computer Science student at Georgia Tech specializing in Computational Media with AI and Game Design. Experienced in cybersecurity, machine learning, and full-stack development with a passion for creating innovative solutions that bridge the gap between security and artificial intelligence.
About Me
Beyond coding and cybersecurity, I'm a passionate individual with diverse interests that fuel my creativity and drive. When I'm not building AI models or securing systems, you'll find me exploring the world around me.
🏃♂️ Active Lifestyle
🎨 Creative Pursuits
🌍 Global Perspective
Professional Experience
- Conducted research on Artificial Intelligence, data collection methods, and physics applications to advance STEM education
- Designed a data scraper to collect and analyze datasets for research, improving data accessibility and usability (Selenium)
- Developed a Physics GPT model to generator accurate and interactive physics explanations, enhancing educational resources
- Assisted in planning, promoting, and executing the Computer Science based events at the Atlanta Science Festival
- Developed an automated model to systematically analyze codebases, mitigating companywide risks and vulnerabilities by 40%
- Released a comprehensive Software Assurance Plan defining security standard across Lockheed's aerospace program
- Implemented DISA STIGS framework and conducted security testing, ensuring DoD Cyber Security standard adherence
- Collaborated with cross-functional teams to perform code reviews and compliance audits, identifying and remediating risks
- Conducted research on Artificial Intelligence, data collection methods, and physics applications to advance STEM education
- Designed a data scraper to collect and analyze datasets for research, improving data accessibility and usability (Selenium)
- Developed a Physics GPT model to generator accurate and interactive physics explanations, enhancing educational resources
- Assisted in planning, promoting, and executing the Computer Science based events at the Atlanta Science Festival
- Developed robust security framework for mobile and IoT systems enhancing security using APK files firmware (APKtools)
- Created an automated model which would download multiple APK's from the internet and place into a google sheet (Selenium)
- Analyzed the structure of Arduino's PLC devices to undergo penetration testing to create a more robust model of security
- Created projects which automated over various devices using the Arduino device to ensure speed to everyday tasks
- Presented data-driven insights to the CEO, resulting in 12% conversion increase pet wellness tailored via Shopify campaigns
- Identified purchase patterns from 5000+ customers data across Amazon FBA and Shopify, improving targeted precision
- Utilized the statistical decision tree model that boosted click rates through personalized product recommendations
Featured Projects
PhysicsGPT
Built a transformer model to answer physics questions using a custom dataset of 60,000+, improving engagement and accuracy. Designed and implemented data preprocessing pipelines to clean, structure, and optimized data for chatbot integration. Leveraged NLP techniques to enhance response accuracy, ensuring reliable and detailed answers for complex physics queries.
DataScrapper
Developed a web scraping tool that extracts information based on user prompts, automating data collection from various websites. Integrating Google based software to store and organize scraped data, streamlining accessibility and usability. Utilizing Python libraries (Beautiful Soup, Selenium) to parse filter and structure extracted data efficiently for better reading.
Stock Tracker
Created a custom stock tracker that analyzes market data and my personal trading behavior patterns. Delivers real-time desktop notifications, helping me identify optimal buy and sell points aligned with past actions. The tool improved my discipline by highlighting emotional trades and reinforcing decisions grounded in historical performance data.
GeNN CI/CD Workflow Integration
Designed and implemented six custom GitHub Action workflows to automate CI/CD testing for GeNN Neural Network framework. Improved pull request stability through the addition of linting (clang-format) and unit-testing, reducing manual review by 40%. Validated workflows on Ubuntu runners, debugging EC2 compatibility refactoring jobs for cross-platform deployment.