Seuring AI Applications with Keycloak and Envoy Gateway in Kubernetes
Last Updated on November 2, 2026 by Editorial Team
Author(s): Yogender Pal
Originally published on Towards AI.
1. The architecture
AI applications need more than a connection to a large language model. When multiple users interact with an AI service, the application must establish who they are and determine which operations they are permitted to perform.
The article walks through a hands-on Kubernetes setup that secures a locally deployed AI chat backend using Keycloak for authentication (issuing OpenID Connect tokens) and Envoy Gateway for JWT validation and role-based authorization before requests reach a FastAPI service backed by Google’s Gemini API. It details the overall architecture and components, how to create a local kind cluster, install Keycloak/Postgres and configure realms/roles and users, build and deploy the FastAPI app, and configure Envoy Gateway routes and a SecurityPolicy that enforces CORS, denies by default, and allows specific paths only for users with the right Keycloak realm roles. It then tests authorization behavior with curl requests (verifying access for ai-user vs ai-admin), adds a Vite frontend that must use the correct Keycloak client/audience settings, explains why “admin” login may fail due to authentication vs authorization and token audience, and hardens the system further with additional checks (audience validation, clientId constraints) and a Kubernetes NetworkPolicy that only allows ingress to the FastAPI pods from the Envoy gateway on the backend port—preventing direct bypasses within the cluster.
Read the full blog for free on Medium.
Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.
Published via Towards AI
Towards AI Academy
We Build Enterprise-Grade AI. We'll Teach You to Master It Too.
15 engineers. 100,000+ students. Towards AI Academy teaches what actually survives production.
Start free — no commitment:
→ AI Engineering For Production — our new book, coming before the end of 2026. Join the waitlist.
→ Agents Architecture Cheatsheet — 3 years of architecture decisions in 6 pages
Our courses:
→ AI Engineering Certification — 90+ lessons from project selection to deployed product. The most comprehensive practical LLM course out there.
→ Agent Engineering Course — Hands on with production agent architectures, memory, routing, and eval frameworks — built from real enterprise engagements.
→ AI for Work — Understand, evaluate, and apply AI for complex work tasks.
Note: Article content contains the views of the contributing authors and not Towards AI.
Related posts
Recent Posts
![[Day 37/100] Build Your Own Agent Memory Layer with Postgres and pgvector](https://miro.medium.com/v2/resize:fit:700/1*-Doos2nmUT3FPZ-Qxbk3Hw.png "[Day 37/100] Build Your Own Agent Memory Layer with Postgres and pgvector")
