Manish Pandey Manish Pandey Cybersecurity Architect & Researcher

cat ./talks/archive

Talk archive.

Archive of talks, workshops, and briefings on cybersecurity, threat modeling, adversary simulation, and AI/ML security.

Speaker 2024

AI/ML Application Red Teaming

Maharshi Markandeshwar University

Practitioner walk-through of red-teaming methodology for LLM-powered applications and ML systems, with reproducible exercises.

AI SecurityRed TeamingLLM
Key takeaways
  • Red teaming LLMs is not penetration testing with a new target
  • Reproducible adversarial test design beats one-off prompt tricks
  • The EchoLeak class of vulnerabilities sits between classical AppSec and ML safety
Speaker 2024

Closed-door briefing on AI/ML security

Industry roundtable (private)

Invited briefing for senior security leaders on threat modeling for agentic AI, the gaps between framework guidance and production deployments, and concrete remediation patterns.

AI SecurityThreat ModelingExecutive Briefing
Key takeaways
  • Framework guidance (STRIDE, ASI Top 10) is a starting point, not a checklist
  • Agentic AI introduces coordination failure modes no single-model threat model catches
  • Remediation patterns from classical security transfer; the gap is in scoping
Interviewee 2024

AI/ML security: interview with a venture capital fund

VC investor briefing

Conversation on the state of AI security tooling, where the market is genuinely under-served, and how technical due diligence should change for AI-native companies.

AI SecurityMarket AnalysisVC
Key takeaways
  • AI security tooling is 3–5 years behind the threat landscape
  • Technical due diligence for AI-native companies needs a security lens, not just a model-evaluation lens
  • The most under-served segment is inference-time protection for production LLM deployments
Host 2023–2024

Vapra.shiksha: hosted talk series

Vapra.shiksha

Hosted a series of talks on cybersecurity craft and applied AI security for an engineering audience.

CommunityEducationAI Security
Key takeaways
  • Practitioner content beats theoretical frameworks for engineering audiences
  • The best talks leave the audience with a tool, not a concept

← Speaking · Book a talk

esc

Type to search. ↑↓ to navigate, ↵ to open.