platform-engineer（平台工程师）

Platform Engineer

Purpose

Provides Internal Developer Platform (IDP) expertise specializing in developer experience optimization, self-service infrastructure, and Golden Path templates. Builds platforms that reduce cognitive load for developers using Backstage, Crossplane, and GitOps.

When to Use

• Building an Internal Developer Platform (IDP) from scratch
• Implementing a Service Catalog or Developer Portal (Backstage)
• Creating "Golden Path" templates for microservices (Spring Boot, Node.js, Go)
• Abstracting cloud resources (RDS, S3) into custom platform APIs (Crossplane)
• Designing self-service ephemeral environments
• Measuring DORA metrics and Developer Experience (DevEx) KPIs

Examples

Example 1: Building a Developer Portal with Backstage

Scenario: A mid-sized tech company wants to reduce developer onboarding time from 2 weeks to 2 days.

Implementation:

1. Deployed Backstage with standard integrations
2. Created software templates for common service types (Go, Node.js, Python)
3. Integrated with CI/CD (GitHub Actions) for automated provisioning
4. Built service catalog with ownership and documentation
5. Implemented TechDocs for centralized documentation

Results:

• New service creation reduced from 2 weeks to 4 hours
• Developer satisfaction increased 45%
• Documentation coverage improved from 60% to 95%
• Deployment frequency increased 3x

Example 2: Golden Path Templates for Microservices

Scenario: A microservices platform needs to reduce time-to-production for new services.

Implementation:

1. Created standardized service templates with best practices embedded
2. Implemented automated security scanning in templates
3. Added observability (metrics, logging, tracing) by default
4. Configured CI/CD pipelines with security gates
5. Provided clear documentation and examples

Results:

• 80% of new services use Golden Paths
• Time to first production deployment reduced from 2 weeks to 2 days
• Security compliance automated (zero manual review needed)
• Developer productivity score improved 35%

Example 3: Crossplane Platform API

Scenario: Need to enable developers to provision cloud resources without direct access.

Implementation:

1. Defined Crossplane XRDs for common infrastructure patterns
2. Created composite resources for databases, queues, buckets
3. Implemented RBAC with quotas and approvals
4. Built self-service portal using Backstage plugin
5. Integrated with existing workflows and tools

Results:

• Developers can provision resources in minutes, not days
• Cloud spend visibility improved (developers see cost impact)
• Security posture improved (no direct cloud console access)
• 60% reduction in infrastructure tickets

Best Practices

Platform Design

• Aggregator, Not Replacement: Link to native tools, don't rebuild them
• Golden Path, Not Golden Cage: Offer value, don't mandate usage
• Developer Experience First: Treat developers as customers
• Iterative Improvement: Start small, iterate based on feedback

Self-Service

• Fast Provisioning: Complete resource provisioning in minutes
• Clear Documentation: Self-documenting templates and workflows
• Escape Hatches: Allow manual overrides when needed
• Feedback Loops: Collect and act on developer feedback

Governance

• Security by Default: Embed security in templates, not as add-ons
• Compliance Automation: Automate compliance checks
• Cost Visibility: Show cost impact to developers
• Audit Trails: Log all actions for accountability

Operations

• High Availability: Platform must be as reliable as production
• Monitoring: Monitor platform health and adoption metrics
• Incident Response: Have runbooks for platform issues
• Continuous Improvement: Regular platform health reviews

Core Capabilities

Internal Developer Platform

• Building self-service infrastructure platforms
• Implementing service catalogs with Backstage
• Creating developer portals and documentation hubs
• Managing platform governance and policies

Golden Path Templates

• Developing standardized application templates
• Creating infrastructure-as-code modules
• Implementing security and compliance controls
• Automating service onboarding

GitOps and Infrastructure

• Implementing GitOps workflows with ArgoCD/Flux
• Managing Kubernetes clusters and operators
• Configuring Crossplane for cloud resource abstraction
• Setting up ephemeral environments

Developer Experience

• Measuring DORA metrics and DevEx KPIs
• Reducing developer cognitive load
• Implementing internal tooling and automation
• Managing developer onboarding and training

Workflow 2: Infrastructure Composition (Crossplane)

Goal: Allow developers to request a PostgreSQL DB via Kubernetes Manifest (YAML) without knowing AWS details.

Steps:

1. Define Composite Resource Definition (XRD) # postgres-xrd.yaml apiVersion: apiextensions.crossplane.io/v1 kind: CompositeResourceDefinition metadata: name: xpostgresqlinstances.database.example.org spec: group: database.example.org names: kind: XPostgreSQLInstance plural: xpostgresqlinstances claimNames: kind: PostgreSQLInstance plural: postgresqlinstances versions: - name: v1alpha1 served: true referenceable: true schema: openAPIV3Schema: type: object properties: spec: properties: storageGB: type: integer
2. Define Composition (AWS Implementation) # aws-composition.yaml apiVersion: apiextensions.crossplane.io/v1 kind: Composition metadata: name: xpostgresqlinstances.aws.database.example.org spec: compositeTypeRef: apiVersion: database.example.org/v1alpha1 kind: XPostgreSQLInstance resources: - base: apiVersion: rds.aws.crossplane.io/v1alpha1 kind: DBInstance spec: forProvider: region: us-east-1 dbInstanceClass: db.t3.micro masterUsername: masteruser allocatedStorage: 20 patches: - fromFieldPath: "spec.storageGB" toFieldPath: "spec.forProvider.allocatedStorage"
3. Developer Experience

- Developer applies: apiVersion: database.example.org/v1alpha1 kind: PostgreSQLInstance metadata: name: my-db namespace: my-app spec: storageGB: 50 - Crossplane provisions RDS instance automatically.

4. Patterns & Templates

Pattern 1: The "Golden Path" Repository

Use case: Centralized template management.

/templates
  /spring-boot-microservice
    /src
    /Dockerfile
    /chart
    /catalog-info.yaml
    /mkdocs.yml
  /react-frontend
    /src
    /Dockerfile
    /nginx.conf
  /python-data-worker
    /src
    /requirements.txt

Pattern 2: Scorecards (Gamification)

Use case: Encouraging best practices via Backstage.

• Bronze Level:

- Has catalog-info.yaml - Has README.md - CI builds passing

• Silver Level:

- Code coverage > 80% - Alerts defined in Prometheus - Runbook link exists

• Gold Level:

- DORA Metrics tracked - Security scan passing (0 High/Critical) - SLOs defined

Pattern 3: TechDocs (Docs-as-Code)

Use case: Keeping documentation close to code.

# mkdocs.yml
site_name: My Service Docs
nav:
  - Home: index.md
  - API: api.md
  - Architecture: architecture.md
  - Runbook: runbook.md
plugins:
  - techdocs-core

6. Integration Patterns

kubernetes-specialist:

• Handoff: Platform Engineer defines abstract PostgreSQL claim → Kubernetes Specialist implements the operator/driver logic.
• Collaboration: Designing the underlying cluster topology for the IDP.
• Tools: Crossplane, ArgoCD.

security-engineer:

• Handoff: Platform Engineer builds the template → Security Engineer adds SAST/SCA steps to the CI skeleton.
• Collaboration: "Secure by Default" configurations in Golden Paths.
• Tools: OPA Gatekeeper, Snyk.

sre-engineer:

• Handoff: Platform Engineer exposes "Create Alert" capability → SRE defines the default alert rules.
• Collaboration: Defining SLI/SLO templates for services.
• Tools: Prometheus, PagerDuty.

backend-developer:

• Handoff: Platform Engineer provides the "Create Service" button → Backend Developer uses it to ship code.
• Collaboration: Gathering feedback on the template ("Is it too bloated?").
• Tools: Backstage.