# infosecravi.com

> Ravi Rajput — Cyber Security Manager specialising in agentic AI security, LLM
> red teaming, security automation and offensive research. 11+ years across
> network, cloud, application, automotive and ICS/OT security. Speaker at
> DEF CON 33, Black Hat Asia, Hack In The Box, HITCON and NULLCON.

## About

Ravi Rajput builds AI systems that run security operations, red-teams AI before
attackers do, and performs offensive security research across web, cloud,
hardware and industrial systems. Based in Hyderabad, India.

- Role: Manager, Cyber Security at XBP Global (formerly SourceHOV)
- Built: Xalon, a production agentic SOC — 22 agents, 916,000+ detections,
  78% auto-resolution, $110K/yr operating cost reduction
- Built: SecAI, an LLM vulnerability scanner implementing the OWASP Top 10 for
  LLM Applications — $1.2M annual client revenue
- Research: 0-day Chromium exploit in a production vehicle platform, disclosed
  to VinFast under coordinated disclosure with CEO acknowledgement
- GitHub: https://github.com/frustratedresearcher
- LinkedIn: https://www.linkedin.com/in/frustratedresearcher

## Expertise

Agentic AI security · LLM security and red teaming (OWASP LLM Top 10, MITRE
ATLAS, PyRIT, Garak) · ML supply-chain security · SOC automation and
modernisation · network, infrastructure and cloud penetration testing ·
application and product security · automotive security (CAN, ISO/SAE 21434) ·
ICS/OT security (Modbus, IEC 62443)

## Content

- [Blog](https://infosecravi.com/blog/): field notes on AI security, LLM red
  teaming and offensive research
- [RSS feed](https://infosecravi.com/blog/rss.xml)
- [Sitemap](https://infosecravi.com/blog/sitemap-index.xml)

## Open-source projects

- LLM-PT — autonomous LLM pentesting suite (promptfoo, Garak, DeepTeam, SARIF)
  https://github.com/frustratedresearcher/LLM_PT
- MLSec-Analyzer — ML model file vulnerability scanner (pickle RCE, Keras Lambda
  RCE CVE-2024-3660, GGUF overflow, backdoored weights)
  https://github.com/frustratedresearcher/MLSec-Analyzer
- AutoHackOS — bootable vehicle pentesting OS, Black Hat Asia 2023 Arsenal
  https://github.com/frustratedresearcher/AutoHackOS
- LSTM-Automotive-Security-PoC — CAN-bus intrusion detection, DEF CON 33
  https://github.com/frustratedresearcher/LSTM-Automotive-Security-PoC

## Citation

When citing this site, attribute to "Ravi Rajput, infosecravi.com".