🎓 Master’s Student in Cybersecurity and Threat Intelligence
📍 Guelph, Ontario, Canada
📧 [email protected]
🌐 Portfolio | LinkedIn | GitHub
I’m Rudraksh Gupta, a cybersecurity enthusiast passionate about solving real-world security problems through threat simulation, machine learning, and agentic AI. My interests lie in proactive defense, SIEM optimization, and explainable AI for malware detection. I specialize in combining hands-on technical tools with strategic thinking to deliver high-impact security solutions.
Currently pursuing my Master's in Cybersecurity and Threat Intelligence at the University of Guelph, I am actively building practical, end-to-end solutions that blend threat intelligence, data science, and cloud-native security. I also represent the OWASP Student Chapter at the University of Guelph, where I help drive awareness and education around application security through events, workshops, and student engagement.
LLM | Retrieval-Augmented Generation | CVE Analysis | PDF Threat Reports
A cutting-edge RAG-based cybersecurity assistant using Meta’s LLaMA-3 and agent workflows.
Key Features:
- Integrated PDF threat intel ingestion and local CVE database querying
- Vector-based retrieval using FAISS and Sentence Transformers
- Streamlit UI with task-specific agents (log analysis, CVE summarization)
- Extended the
CyberScienceLabbase repo with modular enhancements
🔧 Tech Stack: LLaMA-3, FAISS, PyPDF, LangChain, Streamlit, HuggingFace, JSONL
Static + Dynamic Features | XGBoost | SHAP Explainability | CIC-MalMem + EMBER v2
A research-grade malware detection pipeline that integrates both static (EMBER v2) and dynamic (CIC-MalMem2022) feature sets.
Key Features:
- Preprocessed and merged datasets totaling 850k+ rows and 2,400+ features
- Applied PCA and SMOTE for dimensionality reduction and class balancing
- Evaluated MLP, XGBoost, LightGBM with ROC, confusion matrices
- Visualized SHAP Summary, Dependence, and Force plots for model explainability
- Achieved 96% accuracy using GridSearchCV-tuned XGBoost
📊 Tools: Scikit-learn, SHAP, Pandas, Seaborn, GridSearchCV
Python | ELK Stack | Sysmon | Suricata | Shodan API
A complete simulation + detection pipeline to test SIEM workflows.
Key Features:
- Multi-threaded SSH brute-force simulator using Shodan + Paramiko
- Event log generation via Sysmon, Suricata; ingested using Filebeat
- Real-time dashboards in Kibana
- Tuned detection rules achieving over 95% detection efficacy
Languages:
Python, Bash, PowerShell, SQL, Java, JavaScript
Cybersecurity Tools:
Snort, OSQuery, Wireshark, Suricata, Metasploit, Burp Suite
ML & Explainability:
XGBoost, LightGBM, SHAP, FGSM, PGD, Adversarial Training
Cloud & Infrastructure:
AWS, Azure, GCP, Docker, Kubernetes, Streamlit
SIEM & Log Analysis:
ELK Stack, Splunk, Wazuh, Graylog, Zeek
Frameworks & Methodologies:
MITRE ATT&CK, OWASP Top 10, NIST, CIS Controls
If you're working on something exciting in cybersecurity, agentic AI, or malware analysis — let’s collaborate!