AI MLOps Pipeline
AI MLOps Pipeline是一款learning方向的AI技能，核心价值是Multi-agent orchestrated approach to building production ML pipelines covering the entire lifecycle: data engineering, feature engineering, model training, Kubernetes deployment, and observability，可用于解决开发者在learning领域的实际问题，帮助用户提升效率、自动化重复任务或优化工作流。
Multi-agent orchestrated approach to building production ML pipelines covering the entire lifecycle: data engineering, feature engineering, model training, Kubernetes deployment, and observability.
mkdir -p ./skills/ai-mlops-pipeline && curl -sfL https://raw.githubusercontent.com/mayurrathi/awesome-agent-skills/main/skills/ai-mlops-pipeline/SKILL.md -o ./skills/ai-mlops-pipeline/SKILL.mdRun in terminal / PowerShell. Requires curl (Unix) or PowerShell 5+ (Windows).
Skill Content
# AI MLOps Pipeline
Purpose
Build a complete MLOps pipeline from data to production with end-to-end orchestration.
Phase 1: Data & Requirements
- Source audit and schema validation
- Data versioning (DVC, lakeFS)
- Incremental loading with CDC
- Quality gates and anomaly detection
- Bronze/Silver/Gold lakehouse architecture
Phase 2: Feature Engineering & Training
- Feature store (Feast, Tecton)
- Hyperparameter optimization (Optuna, Ray Tune)
- Distributed training (Horovod, PyTorch DDP)
- Experiment tracking (MLflow, W&B)
- Model registry with versioning
Phase 3: Production Deployment
- REST/gRPC serving (FastAPI, BentoML, TorchServe)
- Batch inference via Airflow/Kubeflow
- Deployment: Blue-green, Canary, Shadow, A/B
- Kubernetes: GPU allocation, auto-scaling, Istio
Phase 4: Monitoring
- Drift detection (KS test, PSI)
- Performance tracking (latency, throughput)
- Automated retraining on drift
- Rollback within 5 minutes
- Observability: Prometheus, Grafana, Jaeger
Configuration
Options: MLflow/W&B, Feast/Tecton, KServe/Seldon, Airflow/Prefect, AWS/Azure/GCP
🎯 Best For
- UI designers
- Product designers
- Claude users
- ChatGPT users
- Gemini users
💡 Use Cases
- Generating component mockups
- Creating design system tokens
- Using AI MLOps Pipeline in daily workflow
- Automating repetitive learning tasks
📖 How to Use This Skill
- 1
Install the Skill
Copy the install command from the Terminal tab and run it. The SKILL.md file downloads to your local skills directory.
- 2
Load into Your AI Assistant
Open Claude or ChatGPT and reference the skill. Paste the SKILL.md content or use the system prompt tab.
- 3
Apply AI MLOps Pipeline to Your Work
Provide context for your task — paste source material, describe your audience, or share existing work to guide the AI.
- 4
Review and Refine
Edit the AI output for accuracy, tone, and completeness. Add human insight where the AI lacks context.
❓ Frequently Asked Questions
Does this work with Figma?
Some design skills integrate with Figma plugins. Check the Works With section for supported tools.
How do I install AI MLOps Pipeline?
Copy the install command from the Terminal tab and run it. The skill downloads to ./skills/ai-mlops-pipeline/SKILL.md, ready to use.
Can I customize this skill for my team?
Absolutely. Edit the SKILL.md file to add team-specific instructions, examples, or workflows.
⚠️ Common Mistakes to Avoid
Skipping usability testing
AI-generated designs should be validated with real users before development.
Not reading the full skill
Skills contain important context and edge cases beyond the quick start.