_______________________________________________________________________________________________
/ "Breaking systems in controlled labs to understand how they tick; engineering AI guardrails \
\ to defend critical environments against modern social engineering and adversarial vectors." /
-----------------------------------------------------------------------------------------------
\ ^__^
\ (oo)\_______
(__)\ )\/\
||----w |
|| ||
[SYSTEM_LOG: PROFILE_INITIALIZED]
> OPERATOR : Umar Farooque (@UmarFarooqueJi)
> ROLE : Cybersecurity Researcher & AI Defense Engineer
> CORE_FOCUS : Application Security (OWASP Top 10), Adversarial LLM Security, Threat Detection
> FLAGSHIP : SafeSignal — Autonomous Multi-Tier AI Scam Defense Engine
> LAB_STACK : Kali Linux | Python | Dart/Flutter | Supabase (RLS) | Firebase | VS Code | Antigravity
I'm Umar Farooque, an offensive and defensive security researcher based in India. My work centers on Adversarial AI Security, Application Security Hardening, and Autonomous Workflow Automation.
I focus on practical engineering: building resilient defenses by analyzing real attack vectors:
- 🛡️ AppSec & Penetration Testing: Rigorous assessment across Web Applications (OWASP Top 10), Mobile Testing (Android APK decompilation & Manifest auditing), and Network Footprinting.
- 🤖 Adversarial AI & LLM Hardening: Identifying and mitigating prompt injection attack vectors (Direct & Indirect), system prompt leakage, jailbreak containment, and RAG retrieval poisoning.
- 🚀 Flagship Engineering: Creator of SafeSignal — an open-source, multi-tier agentic engine intercepting deceptive voice calls, phishing SMS, and fraud screenshots with sub-second AI triage.
- ⚙️ Automated Security Tooling: Crafting high-throughput automation pipelines with Python, Bash, hardened REST APIs, and modern datastores (Supabase PostgreSQL with RLS, MongoDB).
🛡️ [01] OFFENSIVE & DEFENSIVE CYBERSECURITY
| Area | Focus Concepts & Vulnerability Testing | Toolchain & Utilities |
|---|---|---|
| Web App Security | SQL Injection (Manual & Automated), XSS (Stored/Reflected/DOM), CSRF, SSRF, IDOR, SSTI, HTTP Smuggling, LFI/RFI | Burp Suite, OWASP ZAP, Sqlmap, Nikto, Dirb |
| Mobile App Testing | Android architecture, APK decompilation, Manifest auditing, static & dynamic security analysis | MobSF, Apktool, Drozer |
| Network & Footprinting | Active/Passive Recon, OSINT, Port scanning, Traffic sniffing, MITM analysis, Protocol troubleshooting | Wireshark, Nmap, Kali Linux, Cisco Packet Tracer |
| System & Investigation | Password testing, Linux system hardening, Log analysis, Disk & Memory evidence acquisition | Metasploit, Hydra, John the Ripper, Autopsy |
🧠 [02] ADVERSARIAL AI & AUTOMATION ENGINEERING
| Topic | Focus & Research Labs | Frameworks & Stacks |
|---|---|---|
| AI Security Testing | Direct & Indirect Prompt Injection defense, Jailbreak payload stress-testing, System prompt containment | Google Gemini, LLM Guardrails |
| RAG & Vector Grounding | Vector embeddings, semantic similarity search, scam knowledge base retrieval pipelines | Vector DBs, LangChain concepts, Embeddings |
| Workflow Automation | Multi-agent execution pipelines, automated developer tooling, security workflow scripting | Antigravity AI, Python CLI, Bash Scripts |
☁️ [03] BACKEND, CLOUD & DEVELOPMENT ENVIRONMENT
| Layer | Technologies & Implementations |
|---|---|
| Cloud Security | AWS Fundamentals (IAM least privilege, VPC network segmentation, S3 bucket policies, CloudTrail auditing) |
| Databases | Supabase (PostgreSQL with Row Level Security), MongoDB (NoSQL datastores), Firebase (Firestore & Auth) |
| Application Layer | Python (Automation scripts), Dart & Flutter (Cross-platform production apps), REST APIs (JWT/Bearer Auth, Webhooks) |
| Workspace & Tooling | VS Code, Antigravity (Agentic AI IDE), Git & GitHub (Version Control & CI/CD Pipelines) |
[SYSTEM_ARCHITECTURE: SAFESIGNAL]
├── 📥 Input Telemetry -> Suspicious Calls, Phishing SMS, Fraudulent Screenshots
├── ⚡ High-Speed Triage -> Signature checks, phone fraud database hashing (<300ms)
├── 🧠 Multi-Tier AI Reasoning -> Deep contextual LLM assessment to detect social engineering
└── 🛡️ Vernacular Explanations -> Plain-language actionable safety advice (Hindi/English)
👉 Explore the open-source repository: UmarFarooqueJi/SafeSignal
[OPERATOR: UMAR FAROOQUE] — Open to security research collaborations, AI defense experiments & offensive security inquiries.