Amol Bhure

AI Security Leader · Security Architect

Amol Bhure

I work as an AI security architect, a security leader for 18+ engineers, a red teamer and researcher, and a builder of secure AI agents

13+ years securing products, platforms and people. Today I design secure-by-design AI systems: LLM threat models, AI red teaming, guardrails for agentic workflows, and security programmes that let teams ship AI safely.

  • AI / LLM security
  • Agentic AI & MCP
  • AI red teaming
  • Security architecture
  • Product & cloud security
  • GRC
Connect on LinkedIn See my journey

Impact

Numbers from the work

13+years in security
18+engineers led across AppSec, Cloud, Red Team, SOC and GRC
31businesses in one group, secured under one security operating model
200+repositories with automated SAST, DAST and SCA gates
~60%faster incident response from an MCP-based security automation platform
<14 dayscritical-fix time, down from 45 days
10,000+employees reached by security awareness
200+web, mobile and IoT products penetration-tested

AI security

Every AI threat meets a control

I threat-model LLM applications, RAG pipelines and AI agents, then design the controls that stop each risk, mapped to the OWASP Top 10 for LLM Applications, MITRE ATLAS and the NIST AI RMF.

Prompt injectionInstruction isolation and input validation
JailbreaksSafety evaluations and guardrails
Tool misuse by agentsScoped permissions and human approval
Sensitive-data leakageData controls and exfiltration protection
Retrieval poisoningSource trust and provenance checks
Model extraction & inversionAccess limits and anomaly monitoring
Excessive agent permissionsLeast-privilege agent identity

Expertise

What I bring to a security team

AI, GenAI & agentic security

LLM threat modelling, AI red teaming, RAG and agent security, secure tool calling, MCP, guardrails, AI safety evaluations and AI governance.

Enterprise security architecture

Secure-by-design reference patterns, architecture reviews, trust boundaries, Zero Trust, IAM, encryption and risk-based design decisions.

Product, application & DevSecOps

Secure SDLC, SAST, DAST, SCA, secure code review, penetration testing, CI/CD gates, API security and supply-chain risk.

Cloud, identity & data

AWS, Azure and GCP, IAM and PAM, SSO and MFA, CSPM, WAF, Kubernetes and container security, secrets and data protection.

SecOps, risk & compliance

SOC, SIEM, detection engineering, incident response, vulnerability management, ISO 27001, SOC 2, PCI DSS, GDPR, RBI and CERT-In.

Leadership & strategy

Built and scaled security teams, security roadmaps and operating models, executive risk reporting, and partnership with Product and Engineering.

Journey

The scope keeps growing

Every few years, the thing I protect gets bigger: from a single codebase to AI agents that act on their own. Titles and companies are on LinkedIn; this is the craft.

Arcs rising from a base line, each wider than the last: code at the centre, then web and networks, devices, cloud and products, scale, people and, as the widest arc, AI systems.2013 · Web & networks2015 · Devices2020 · Cloud & products2021 · Scale2024 · People2025 · AI systems2012 · Code
  1. 2012

    Code

    I wrote software first, so I know how it is built and how it breaks.

  2. 2013

    Web & networks

    Broke into web apps and networks for clients across Europe, Asia and the Middle East.

  3. 2015

    Devices

    Took apart mobile apps, firmware and radios across 200+ products: Android, iOS, BLE and Zigbee.

  4. 2020

    Cloud & products

    Designed threat models, cloud controls and card-data encryption for regulated products.

  5. 2021

    Scale

    Turned security into a system: gates in 200+ repositories, critical fixes in under 14 days.

  6. 2024

    People

    Grew the teams behind it: 18+ engineers across AppSec, Cloud, Red Team, SOC and GRC.

  7. 2025

    AI systems

    Now securing LLMs and AI agents: threat models, red teaming, guardrails and MCP.

Certifications

Verified credentials

CCAR-PClaude Certified Architect – ProfessionalAnthropic · Oct 2026Verify on Credly PMLEProfessional Machine Learning EngineerGoogle Cloud · Oct 2026Verify on Credly CCAR-FClaude Certified Architect – FoundationsAnthropic · Sep 2026Verify on Credly CCAO-FClaude Certified Associate – FoundationsAnthropic · Sep 2026Verify on Credly CISMCertified Information Security ManagerISACA · Apr 2026Verify on Credly

Open source

Giving back

Claude Certification Zero to Hero banner

Claude Certification: Zero to Hero

A free, unofficial study guide I wrote after earning three Claude certifications: full tracks for CCAO-F, CCAR-F and CCAR-P, 410 original practice questions, mock exams in the browser, 20 labs that run offline, five real-world projects and a dedicated AI security track.

Open the website View on GitHub

Contact

Let's talk AI security

LinkedInin/amolbhure GitHub@codedevloper Credlyverified badges

Based in Bengaluru, India.