Contact

01 / Personal Website

John "Jack" Pippett

I design secure systems for AI, telemetry, and the teams that depend on both.

Senior security engineer in Chandler, Arizona, aiming at AI security and application security architecture roles. Currently leading enterprise security data platform work, building independent AI security tools, and writing about how modern systems fail in the real world.

Explore work View resume
What can I help with? AI security engineering, secure agent infrastructure, and architecture reviews.
Best-fit work AI security engineering · application security architecture · security data platforms
Projects LLMScout.fit · iscooked.com
Frameworks I've worked with... OWASP LLM Top 10 · NIST AI RMF · MITRE ATLAS · prompt-injection research

Blog, resume, and project notes for security engineering work that is practical, opinionated, and built close to the metal.

02 / Projects

Small products, real security problems.

01

An AI hardware compatibility platform for finding which local models can run safely and effectively on real hardware.

Visit project 02
iscooked

A security-focused scanner for self-hosted LLM deployments, unsafe local AI exposure, and risky environment configurations.

Visit project

03 / Resume

A resume built around systems that actually run.

Centene Corporation

Threat Infrastructure Senior Cybersecurity Engineer

Own architecture and delivery for Centene's Databricks-based AWS Security Data Lake and Databahn rollout, translating executive goals and vendor patterns into working endpoint telemetry systems across RBAC, disaster recovery, ingestion quality, and downstream analytics.

  • Drove project planning, architecture decisions, stakeholder alignment, and delivery sequencing for enterprise telemetry work spanning security engineering, vendor teams, and platform owners.
  • Led Databahn endpoint telemetry implementation for 70,000+ endpoints, compressing rollout timelines while preserving ingestion quality, access control, and recovery design.
  • Built the automation spine behind the program: GitLab CI/CD promotion gates, Tanium REST API workflows, Sysmon deployment, access reviews, and JIRA remediation pipelines.
M.S. Cybersecurity Capstone

Self-hosted AI agent infrastructure security research

Graduate capstone focused on the security architecture of local AI agent stacks: Ollama/LM Studio style runtimes, shell and environment access, tool permissions, prompt boundaries, and control-plane exposure.

Mapped the work against OWASP LLM Top 10, NIST AI RMF, and MITRE ATLAS, and independently identified indirect prompt injection through environment-variable exposure as a practical attack path.

Education
Western Governors University

M.S. Cybersecurity and Information Assurance

Completed 2026. Capstone centered on AI agent infrastructure security.

Arizona State University

B.S. Economics

Completed 2020. Quantitative background before moving fully into cybersecurity engineering.

Certifications
CompTIA PenTest+ CySA+ Network+
01

AI security architecture

OWASP LLM Top 10, NIST AI RMF, MITRE ATLAS, agent permissions, indirect prompt injection, local LLM exposure.

02

Application security

Threat modeling, secure design review, vulnerability remediation workflows, SDLC guardrails, pragmatic risk framing.

03

Security data platforms

Databricks, Databahn, endpoint telemetry, normalization, enrichment, RBAC, disaster recovery, downstream analytics.

04

Security automation

Python, REST APIs, GitLab CI/CD, Tanium automation, access review tooling, JIRA remediation pipelines.

04 / Blog

A blog for the parts that do not fit on a resume.

Featured blog note

I Gave an LLM Shell Access to a Server. Here's What I Learned.

Spinning up a hardened VPS, handing an LLM shell access, and watching what sticks when AI agents meet real infrastructure.

Read the note

LLM shell access lessons

The real iscooked.com post on agent hardening, prompt injection tests, OS-level controls, and why refusal behavior is not a security boundary.

All blog notes

Currently one real note, one tiny archive, and a suspicious amount of paper texture. More field reports soon.