odock.ai
Blog

Odock Blog: LLM Infrastructure, Security, and Cost Control

Practical articles for teams building AI products with multiple providers, MCP tools, security guardrails, and production governance requirements.

11
All articles
10
Topics
June 11, 2026
Updated
Featured article
AI Security
June 11, 20269 min

AI Security in 2026: Prompt Injection, Tool Poisoning, and the New Agentic Risk Stack

AI security is no longer only about bad prompts. It now includes tool misuse, MCP poisoning, unbounded consumption, and response-side leakage. This post compares those risks with Odock's actual runtime controls.

ai infrastructure securityprompt injectiontool poisoningagentic ai security
YK

Youcef Kaddour

Founder at Odock and AI infrastructure engineer

Read article

What you should take away

  • 1

    The latest AI security guidance is shifting from prompt-only thinking to agent, tool, and runtime control.

  • 2

    Our Security Engine maps well to prompt injection, redaction, leakage, tool governance, and unbounded-consumption controls that belong in the gateway.

  • 3

    Some risks, such as model supply-chain attestation and training-data poisoning, still need controls outside the gateway.

All articles

Clear, technical writing for AI platform teams

Security Architecture
June 11, 2026

How to Build a Lifecycle-Aware AI Security Engine

Prompt safety, tool permissions, budget enforcement, and response leakage do not become visible at the same time. A real AI security engine has to enforce controls in stages.

YK

Youcef Kaddour

8 min
Read article
LLM Infrastructure
April 27, 2026

What Is an LLM Gateway and Why AI Teams Need One Before Production

As soon as AI moves beyond a prototype, teams hit provider sprawl, fragile routing, weak governance, and runaway cost. This article explains the job an LLM gateway actually does and why Odock exists.

YK

Youcef Kaddour

8 min
Read article