GITAM University/CSE/3rd Year/Vizag
Security researcher and AI developer building cloud security tools, LLM systems, and production-ready web applications. I focus on clear evidence, usable interfaces, and reliable engineering.
Technical writing, project reports, and evidence-backed security analysis
I'm a 3rd year Computer Science student at GITAM University, Visakhapatnam — Core CSE, working at the intersection of offensive security and applied AI. I don't just study how systems behave. I study how they break, why they break, and what it takes to make them impossible to break the same way twice.
My interests span cybersecurity, AI & LLMs, RAG systems, security research, and cloud security. My current flagship work is Hollow Purple — a research-grade adaptive cloud identity detection platform that models GCP IAM as live temporal graphs and surfaces multi-stage attack chains with sub-second latency.
Day-to-day: Arch Linux as my daily driver, Kali for offensive work, active bug hunting across Bugcrowd, HackerOne and Intigriti. I pull apart binaries, build AI systems that make real decisions, and led a team to national stage at Smart India Hackathon 2024.
I build AI workflows with clear responsibilities, retrieval context, evaluation checkpoints, and practical user interfaces. The goal is reliable decision support, not flashy demos that collapse under real usage.
My security work prioritizes reproducible findings, responsible disclosure, and fast validation. I investigate attack paths, test assumptions, and document evidence so engineers can act with confidence.
Strong defense depends on signal quality. I connect telemetry, graph modeling, and severity scoring to turn noisy events into prioritized findings that security teams can review quickly.
Security evidence should survive review. I design around hash-chained events, replayable state, and clear audit trails, so every finding can be traced from signal to source.
I focus on IAM governance, event-driven monitoring, attack-path analysis, and response automation for cloud environments where identity is the real perimeter.
A focused research environment for identity attack paths, graph evidence, tamper-evident logs, deterministic replay, and clean operator-facing signals.
Cognitive simulation engine grounded in entropy physics, Bayesian inference, and information theory.
Unified IoT security scanner — 6-phase pipeline reducing manual toolchain orchestration to a single command.
AI-powered legal document analyzer — parses ToS, Privacy Policies, EULAs and surfaces risk scores via Claude API.
Designed Hollow Purple around temporal identity graphs, deterministic replay, and tamper-evident audit chains so cloud privilege movement can be investigated as a timeline, not a pile of logs.
Built AI assistants, RAG pipelines, document analyzers, and agent-style workflows with strong emphasis on traceability, severity scoring, user clarity, and production-minded testing.
Active across vulnerability research, reverse engineering practice, IAM challenge environments, and security tooling, with a bias toward clear evidence and reproducible findings.
Convert cloud activity streams into ranked identity-path evidence: who touched what, which privilege changed, and how the chain can be replayed for review.
Design LLM workflows that explain why something is risky, attach severity, preserve source context, and stay useful to non-specialists under time pressure.
Shape complex backend intelligence into calm, inspectable interfaces with clear states, strong defaults, and direct paths from finding to action.
I can help investigate attack surfaces, model abuse paths, build proof-of-concept detection logic, and document findings in a way engineers can reproduce.
I am strongest where AI needs structure: retrieval, evaluation, scoring, agent tooling, prompt systems, and interfaces that make model output accountable.
I move quickly across frontend, backend, and integration work while keeping the experience legible, accessible, and useful for repeated professional workflows.
Looking for teams that value curiosity, evidence, hands-on building, and the discipline to turn hard technical ideas into tools that survive real use.
Hire Me
Led team through regional selection to compete at the national stage of Smart India Hackathon 2024. Handled full solution design and team leadership.
Smart India HackathonActively identifying and responsibly disclosing security vulnerabilities in live production systems across public and private programs on Bugcrowd, HackerOne, and Intigriti.
Ongoing · Bugcrowd · HackerOne · IntigritiBinary analysis and RE practice — studying compiled code to identify vulnerabilities and dissect techniques for defensive research and CTFs.
OngoingCombining red team offensive tactics with blue team detection engineering to improve overall security posture in real environments.
Security ResearchDeliberately study outside CS — finance, international economics, ML, animation, MCP. Building context that pure engineers miss.
Ongoing
Full list → LinkedIn Certifications ↗
A compact professional timeline covering education, security research, AI builds, certifications, and project milestones.
Computer Science foundation with a focus on AI, cloud platforms, and security.
GCP IAM research using temporal identity graphs and evidence replay.
Led a team through a national-stage product build and technical demo cycle.
Responsible disclosure, vulnerability research, and offensive security practice.
The next phase is practical and focused: build security products that make cloud risk visible, make AI outputs accountable, and turn research into interfaces teams can trust.
Looking for internships and project work in AI, security research, LLM systems, and cloud security. Open to collaborations, research partnerships, and practical engineering problems.
कर्मण्येवाधिकारस्ते मा फलेषु कदाचन।
मा कर्मफलहेतुर्भूर्मा ते सङ्गोऽस्त्वकर्मणि॥