Parth Shukla

Security Research Engineer - AI Security at Harness (2025 - Present)

Bengaluru, India

• Working on Security for AI and AI for Security.
• Securing AI workloads and dogfooding the Harness platform to deploy agentic security infrastructure (Pinaka) at scale.

AI Engineer at Aikido Security (2025)

San Jose, USA

• Built and deployed security AI agents for automated exploit testing, improving prompt logic and performance.
• Created real vulnerable environments from GitHub Advisories to validate and refine exploit capability of AI Agents.

Security Analyst and Builder at Cequence Security (2024)

San Jose, USA

• Performed live API monitoring and conducted real-time threat audits, dramatically enhancing incident detection and response capabilities.
• Spearheaded the creation of advanced test suites for API Security Testing — fully aligned with OWASP API Top 10 and LLM Top 10 standards.
• Found and reported critical security vulnerabilities in customer environments and Fortune 500 companies.

Security Analyst at Cequence Security (2023)

San Jose, USA

• Performed live API monitoring and conducted real-time threat audits, enhancing incident detection and response.
• Spearheaded advanced test suites aligned with OWASP API Top 10 and LLM Top 10 standards.
• Launched Intent Sentinel, an AI firewall for chat using Next.js and GPT-3.5 for intent and trust filtering.

Student Network Technician at Northeastern University (2022)

Boston, USA

• Spearheaded the Invisible Defense Project, deploying Juniper switches and port patches to strengthen campus network security.
• Configured network switches via CLI and optimized wireless infrastructure across campus.
• Began speaking at security conferences, sharing research on offensive security techniques.

Security Analyst at Tata Consultancy Services (2020)

Gandhinagar, India

• Conducted thorough security assessments to identify and address critical vulnerabilities, enhancing system integrity.
• Built a Java-based backend framework with Spring Tool Suite for internal platforms, enabling access for 10,000+ employees.
• Achieved 2nd place in TCS's global Best Ethical Hacker competition.
• Won the Innovation Pride Award in Q1 for groundbreaking security framework improvements.

Pinaka

Subdomain reconnaissance and API attack surface management platform. Automates recon for Red Team operations with real-time discovery and ML insights.

Technologies: AWS Fargate, Serverless Lambda, Harness CI/CD, Agentic AI, Python

PromptGuard

LLM system prompt vulnerability scanner that identifies prompt injection, content leaks, and behavioral override triggers. Integrated with Gemini.

Technologies: OWASP LLM Top 10, Gemini, Python

Intent Sentinel

AI firewall for LLM input classification. Filters malicious intent and trust levels in real-time chat applications.

Technologies: Next.js, GPT-3.5, TypeScript

LLM Red Team Ranger

Continuous multi-turn adversarial testing tool for LLMs. Simulates prompt injection, data leaks, and roleplay abuse with severity classification.

Technologies: Python, LLM APIs, Red Teaming