Dinesh Kumawat
Cybersecurity and Data Analyst with expertise in vulnerability assessment, threat detection, and security analytics. Experienced in identifying system weaknesses and analyzing security data using Python, Burp Suite, Metasploit, Nmap, Excel, and Power BI to support proactive security and risk management.
Core Skills
Education
University of Mumbai
M.Sc. in Information Technology
- Coursework: Computer Vision, Data Science, Natural Language Processing, Machine Learning
University of Mumbai
B.Sc. in Information Technology
- CGPI: 8.0 / 10
- Coursework: Operating Systems, Computer Networks, Database Management Systems, Artificial Intelligence, Security in Computing, Software Engineering, Software Project Management
Certifications
Google Project Management: Specialization
Coursera
Neo4j Graph Data Science Certification
Neo4j
Certified Ethical Hacker
Expired
Publications
A Unified Multi-Task Vision Transformer for AI-Generated Image Detection and Generator Attribution
Proposes a multi-task Vision Transformer framework that jointly performs AI-generated image detection and generator attribution using a shared transformer backbone with task-specific classification heads. The system includes robustness evaluation under image transformations and an occlusion-based explainability mechanism for interpretable predictions.
Blockchain and Decentralized Digital File Management for Transparent and Inclusive Governance in India
Presents a blockchain-enabled Digital File and Records Management System using Hyperledger Fabric and IPFS to improve transparency, accountability, and accessibility in public administration. The framework integrates MeriPehchaan identity access, privacy-aware redaction pipelines, and multilingual dashboards aligned with India's Digital India and RTI governance frameworks.
DOI: 10.22214/ijraset.2025.74242
Experience
Cybersecurity and Data Analyst
5Star Enterprises (ISP), Thakurli (On Contract)
- Reviewed business processes to uncover high-risk operations, resulting in 30% improvement in risk mitigation
- Conducted over 15 detailed security audits identifying critical system vulnerabilities and providing mitigation strategies
- Collaborated cross-functionally with SMEs to implement secure, cost-effective IT improvements
- Created actionable reports and documentation for security upgrades adopted by clients
Bug Bounty and Tester
Freelance
- Discovered and reported 20+ critical vulnerabilities to various SMEs, some with CVSS scores above 7.0
- Delivered structured technical documentation that led to rapid issue remediation and improved dev team workflows
- Utilized tools like Burp Suite and Nmap to simulate real-world attack scenarios and strengthen platform resilience
Featured Projects
AI Image Authenticity Analyzer (VeriVision)
Deep learning system for detecting AI-generated images and identifying their source models.
- Developed a multi-task Vision Transformer based system to detect AI-generated images and identify likely source models such as Stable Diffusion, DALL·E, and Midjourney.
- Implemented deep learning pipelines for image preprocessing, authenticity classification, and generator attribution.
- Built an interactive Streamlit web application enabling real-time image upload, analysis, and visual explanation of model predictions.
- Designed robustness testing features to evaluate model predictions under image compression and resizing.
Wildlife Sound Identifier
Machine learning system for automatic identification of wildlife species from environmental audio recordings.
- Developed an audio classification system that identifies wildlife species from sound recordings using machine learning.
- Extracted Mel-spectrogram features from audio signals and trained classification models for species prediction.
- Built a Streamlit-based interface for uploading audio files, visualizing spectrograms, and displaying prediction results.
- Designed the system to support biodiversity monitoring and bioacoustic research applications.
Election Result Analyzer
Data scraping and visualization platform for analyzing Indian election results across constituencies and states.
- Built a Python-based data pipeline to scrape constituency-wise election results and extract candidate and party information.
- Developed a Streamlit dashboard providing interactive filters, charts, and visual insights for election data analysis.
- Implemented automated screenshot capture and structured data extraction for candidate votes, rankings, and margins.
- Designed visual analytics including party vote share charts, state-wise seat distribution, and constituency-level analysis.
WhatsApp Chat Analytics Dashboard
Interactive Streamlit dashboard for analyzing WhatsApp chat export data.
- Developed a data analysis tool to process WhatsApp chat export files and generate conversation insights.
- Implemented participant activity analysis, keyword filtering, and time-based chat statistics.
- Created interactive charts and dashboards to visualize messaging patterns and user engagement.
- Enabled easy upload and processing of chat files through a user-friendly Streamlit interface.
Anonymous Complaint System
Secure web platform for anonymous grievance reporting and complaint management.
- Developed a web-based platform allowing users to submit complaints anonymously while preserving confidentiality.
- Implemented complaint tracking, committee-based assignment, and administrative dashboards for complaint resolution.
- Designed secure data handling mechanisms and database structures for managing complaint records.
- Built the system using a LAMP-based architecture with PHP backend and MySQL database.