ANOMALY DETECTION
Unsupervised models learn what normal looks like on a network, then flag deviations such as unusual login times, data exfiltration spikes or lateral movement.
Ethical hacking • penetration testing • vulnerability assessment • security research
A practical security portfolio focused on ethical hacking, defensive awareness, and hands-on security labs.
Cybersecurity practitioner with experience across penetration testing, vulnerability assessment, security research, and practical security challenges. Work is performed within defined legal and authorized scope.
Training covering reconnaissance, scanning, exploitation, malware analysis, social engineering and other ethical hacking domains.
12+ rooms across web exploitation, OSINT, Active Directory and reverse engineering paths.
Practical cybersecurity competition focused on ethical hacking, problem solving and teamwork.
Completed a Certified Ethical Hacking Training Program conducted by Denphlink company.
Hands-on projects covering offensive security, network defense, and automation tooling.
A modular OSINT and recon framework that chains subdomain enumeration, port scanning, and banner grabbing into a single automated pipeline with report export.
Identified and exploited SQLi, stored XSS, and broken auth flaws in a deliberately vulnerable web app, then documented full remediation steps in a client-style report.
Built a small AD lab and executed a full attack chain — enumeration, Kerberoasting, and privilege escalation to Domain Admin — using industry-standard tooling.
Deployed a self-hosted SIEM to ingest logs from a small network, wrote detection rules for common attack patterns, and practiced incident triage workflows.
A packet-capture analysis script that flags anomalous traffic patterns (port scans, ARP spoofing signatures) and summarizes findings for quick review.
Documented solutions and methodology for 15+ CTF challenges spanning web exploitation, binary reverse engineering, cryptography, and forensics.
Where machine learning strengthens defense — detecting threats faster, cutting alert noise — and where AI systems themselves become the attack surface.
Unsupervised models learn what normal looks like on a network, then flag deviations such as unusual login times, data exfiltration spikes or lateral movement.
Classifiers trained on static and behavioral features (API calls, opcodes, file entropy) separate benign from malicious samples and group new variants into families.
NLP models score message text, URLs and sender patterns to catch phishing and business-email-compromise that signature rules miss.
ML-assisted IDS profiles traffic flows to spot scans, C2 beaconing and DDoS patterns, complementing signature-based tools like Snort and Suricata.
User and entity behavior analytics builds baselines per account, while ML-based prioritization reduces false positives so analysts focus on real incidents.
Language models help summarize incidents, explain logs, draft detection rules and speed up reporting — always with a human verifying the output.
from sklearn.ensemble import IsolationForest import pandas as pd # flow features: bytes, packets, duration, dst_ports df = pd.read_csv("network_flows.csv") model = IsolationForest(contamination=0.01, random_state=42) model.fit(df) df["score"] = model.decision_function(df) df["suspicious"] = model.predict(df) == -1 # send top outliers to an analyst for review print(df[df.suspicious].nsmallest(10, "score"))
// Reference: OWASP Top 10 for LLM Applications · MITRE ATLAS · NIST AI Risk Management Framework. Concept demo is illustrative; all testing is performed in authorized lab environments.
12 hands-on cybersecurity challenges covering web, Linux, network, forensics, OSINT, crypto, reverse engineering and red-team methodology.
Linux enumeration and privilege basics.
Analyze a captured packet stream and recover the flag.
Discover an exposed endpoint and validate access controls.
Identify and safely demonstrate an SQL injection flaw in a lab.
Crack a weak password hash using offline analysis.
Trace public clues to identify the simulated target.
Escape a restricted shell in a controlled CTF environment.
Find indicators of compromise hidden in security logs.
Investigate memory corruption in a deliberately vulnerable binary.
Audit weak JWT validation and recover the lab flag.
Reverse a safe CTF sample and identify its execution logic.
Chain enumeration, exploitation, and privilege escalation in a sandbox.
Selected practice tracks and challenge areas presented in the same neon-green CTF visual style as the portfolio certificates.
Selected certification and learning material from the portfolio.
Certificate of appreciation received at NIET.
Looking for a penetration tester, security auditor, or someone to run a vulnerability assessment? Available for freelance engagements, bug bounty collaboration, and security roles.
All security work is conducted within defined legal scope and can include a clear written report.