skills.rest
JaviMontano
JaviMontanoCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

aws-architecture-implementation

Implements AI and GenAI architectures on AWS using Bedrock, SageMaker, and OpenSearch.

Updated Jun 28, 2026
Search Tags:#deployment#aws#rag#iac#cloud architecture#bedrock#sagemaker
One-click install
npx skills add https://github.com/JaviMontano/claude-plugins --skill aws-architecture-implementation-javimontano
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: aws-architecture-implementation
Source: https://github.com/JaviMontano/claude-plugins/tree/main/plugins/claude-native-toolkit/skills/aws-architecture-implementation
Command: npx skills add https://github.com/JaviMontano/claude-plugins --skill aws-architecture-implementation-javimontano

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve? Teams that have an approved AWS AI architecture still face the hard work of turning it into running infrastructure: configuring Bedrock Knowledge Bases, SageMaker pipelines, OpenSearch vector stores, security controls, CI/CD, and cost governance. This Skill converts an existing design into step-by-step implementation guides, IaC templates, and deployment patterns. ## Core Features & Use Cases - Service Setup Playbooks: Guided configuration of Bedrock (model access, Guardrails, Knowledge Bases, Agents), SageMaker (training, registry, endpoints, Model Monitor), and OpenSearch Serverless vector search. - IaC Templates: CDK, CloudFormation, and Terraform patterns for RAG stacks, MLOps stacks, agent stacks, and multi-model routing with tagging and environment promotion. - Security & Cost Controls: VPC endpoints, KMS encryption, least-privilege IAM, WAF rules, CloudTrail auditing, budgets, and FinOps dashboards. - Deployment Automation: Blue/Green, canary, shadow, and feature-flag deployment patterns with rollback procedures and CI/CD gates. - Use Case: Given an approved Bedrock RAG design, produce the full implementation guide covering Knowledge Base sync, OpenSearch index mapping, Guardrails, Lambda query handler, monitoring dashboards, and a CodePipeline release flow. ## Quick Start Ask the assistant to implement your approved AWS AI architecture, for example: implement a Bedrock RAG stack with Knowledge Bases, OpenSearch Serverless, Guardrails, and monitoring.

Frequently Asked Questions about aws-architecture-implementation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I implement a Bedrock RAG architecture on AWS?▼

Create an S3 document bucket with KMS encryption, provision an OpenSearch Serverless vector collection, then create a Bedrock Knowledge Base with a chunking strategy and Titan Embeddings. Sync the data source, attach Guardrails, and front queries with a Lambda handler behind API Gateway.

How to set up a SageMaker pipeline from training to production?▼

Build a SageMaker Pipeline covering processing, training, evaluation, and registration into the Model Registry with an approval workflow. Deploy the approved model to a multi-AZ endpoint with auto-scaling, enable data capture, and configure Model Monitor for data and model quality.

What is the difference between aws-architecture-design and aws-architecture-implementation?▼

The design skill produces the target architecture and service tradeoffs before any build work. The implementation skill takes that approved design and executes it: service configuration, IaC templates, security hardening, CI/CD, and monitoring.

What should I do if Bedrock is not available in my AWS region?▼

Use a cross-region inference profile when compliance allows it. If data residency rules prohibit cross-region calls, fall back to a self-hosted model on SageMaker with equivalent security and monitoring controls.

Which deployment pattern should I use for SageMaker model updates?▼

Blue/Green deployment gives zero-downtime cutover with a fast rollback target, while canary deployment shifts 5-10% of traffic first with CloudWatch alarms triggering automatic rollback. Shadow deployment validates a new model against production traffic without user impact.

When should I not use this AWS AI implementation skill?▼

Do not use it to design a new architecture (use aws-architecture-design), to audit an existing deployment (use aws-architecture-audit), or for general non-AI AWS infrastructure such as landing zones and networking (use infrastructure-architecture).

Related Skills

Explore More Software Engineering Skills→
JaviMontanoJaviMontano

metodologia-aws-architecture-design

Design AWS GenAI architectures for scalable, secure AI workloads.

Community
Advanced
Rene-KuhmRene-Kuhm
1

terraform

Audit Terraform infrastructure as code for security gaps, cost inefficiencies, and policy deviations.

Community
Advanced
tomztomz

aws-devops

Automate AWS CI/CD workflows with CodeBuild, CodePipeline, CodeDeploy, and ECR.

Community
Advanced
alicangnllalicangnll

Cloud Mobile Security

Identify and exploit vulnerabilities in Firebase, AWS, Azure, and GCP mobile platforms.

Community
Advanced