# JIT Chapter 86: Traffic Engineering (/docs/deploying-agent-factories-in-the-cloud/86-traffic-engineering/overview)

---
type: Document
title: "JIT Chapter 86: Traffic Engineering"
description: "Gateway API, Envoy Gateway, KEDA autoscaling, and Envoy AI Gateway for production traffic management"
status: stable
order: 8600
generated:
  by: ag-archive/0.1.0
  at: 2026-08-26T00:00:00Z
ksor:
  audience: [public]
  approval:
    by: ag-archive/0.1.0
    at: 2026-08-26T00:00:00Z
trust_tier: unverified
build_id: sha256:c8912c3008ae345a7fdc33ace656531ad74f37c065177057c390c7d759f1e970
source_commit: dad4758bb59919c7b9130444778ade062a2dce3e
ksor_version: 0.0.47
---

Traffic engineering exposes your Kubernetes agent to the world with safe ingress, routing, and autoscaling. You build a `traffic-engineer` skill that covers Gateway API (the Ingress successor), Envoy Gateway, rate limiting, TLS, traffic splitting, and KEDA-based scaling.

---

## Goals

- Understand Gateway API concepts and deploy Envoy Gateway
- Route traffic with HTTPRoute (paths, headers, query matching)
- Apply rate limiting, retries, timeouts, and circuit breaking
- Terminate TLS with CertManager
- Run canary/blue-green splits and autoscale with KEDA
- Manage LLM traffic with Envoy AI Gateway patterns
- Package the patterns into a reusable traffic-engineer skill

---

## Lesson Progression

- **L00**: Build Your Traffic Engineering Skill
- **L01-L03**: Ingress fundamentals → Gateway API concepts
- **L04-L06**: Envoy Gateway setup; routing; rate limiting and resilience
- **L07-L08**: TLS with CertManager; traffic splitting (canary/blue-green)
- **L09-L10**: KEDA autoscaling; resilience patterns
- **L11**: Envoy AI Gateway for LLM traffic
- **L12**: Capstone—production traffic for the Task API; finalize the skill

Each lesson ends with a reflection to test, find gaps, and improve the skill.

---

## Outcome & Method

You finish with production-ready ingress for the Task API: Gateway API + Envoy Gateway, TLS, rate limiting, resilience, canary/blue-green, KEDA autoscaling, and AI Gateway patterns—plus a reusable traffic-engineer skill. The chapter follows the skill-first flow: learn, apply, capstone, finalize.

---

## Prerequisites

- Chapters 79-85 (containerized, deployed, observable service)
- Gateway/Helm familiarity for controller installs