KEHAO CHEN

SOFTWARE ENGINEER · AI APPLICATIONS · TAIPEI

I build AI applications on production instincts.

Developer first. Grounded in Cloud Native, shaped by SRE, and now focused on turning AI into software people can actually use.

Explore my work ↓GitHub
CORE
AAI Applications
PProduct
CCloud Native
RReliability
A
AI APPSCurrent craft
D
DEVELOPMENTPrimary identity
C
CLOUD NATIVEEngineering foundation
S
SREExperience carried forward

FOCUS & FOUNDATION / 01—04

Building forward, without discarding the past.

My path moved from backend development into DevOps, Cloud Native, SRE, and operations—and now back to development, with AI applications as the main focus.

01

AI Applications · Product Engineering

current focus

Turn AI capability into useful software.

Building AI applications around real workflows: clear product intent, grounded context, evaluable outputs, and interfaces that make intelligence practical rather than ornamental.

AI AgentsLLM AppsEvaluation
02

AI-assisted Development · SDLC

delivery leverage

Build with AI, engineer with intent.

Exploring spec-driven and agentic development workflows that compress delivery time while keeping architecture, verification, and engineering judgment explicit.

Spec-drivenAgentsDeveloper Experience
03

Cloud Native · Platform

production depth

A Cloud Native foundation that survives reality.

Years of backend, Kubernetes, platform, and production experience shape how I build today: deployable, observable, secure, and ready for the constraints beyond a demo.

KubernetesCloud NativeGitOps
04

Reliability · Engineering History

earned perspective

Reliability is experience I carry forward.

SRE, observability, incident analysis, and operations are no longer my primary role. They remain a useful lens for designing AI applications people can actually depend on.

OpenTelemetrySREObservability

THE BUILDING MODEL

From possibility to product.

AI changes what software can do. Cloud Native and reliability experience change whether that software can survive outside the prototype.

01
FrameStart with the user workflow
→
02
GroundGive models useful context
→
03
EvaluateMake quality observable
→
04
ShipEngineer the whole application
→

CERTIFICATIONS

Credentials that support the foundation.

CKA

Cloud Native Computing Foundation

Certified Kubernetes Administrator

VERIFIED SKILL
AZ

Microsoft Certified

Azure Solutions Architect Expert

VERIFIED SKILL
AI

Microsoft Certified

Azure AI Engineer Associate

VERIFIED SKILL

ENGINEERING PRINCIPLES

How I make the work hold together.

01

Design for the operator

A system is not finished when it deploys. It is finished when the person on call can understand what it is doing.

02

Automate the evidence

Reliability, security, and compliance should leave useful traces as a by-product of normal operation.

03

Teach the system

Mentoring, shared language, and small repeatable practices create more leverage than heroic debugging.

04

Use AI with judgment

AI can compress the path from intent to software. Architecture, verification, and responsibility still belong to us.

NOW / TAIPEI

Back to building—this time with AI.

After moving from backend development through DevOps, public cloud, SRE, and operations, I have returned to a developer-focused role. My primary work now is AI application development.

I’m especially interested in production-ready AI applications, agents, spec-driven development, and the engineering practices that turn a promising model interaction into a dependable product.

LinkedIn GitHub