Resource Library
192 curated resources — docs, blogs, videos, repos, and practice exams
192 resources
AWS Skill Builder Exam Prep Plan
Free lessons per domain, SimuLearn labs, practice assessments
Official Practice Question Set (20 free questions)
20 free exam-caliber questions from AWS
Official Exam Guide PDF
Every skill, task, and in-scope service for the AIP-C01 exam
Bedrock Workshop
Jupyter notebooks for every Bedrock feature
Bedrock RAG Workshop
RAG-specific deep dive with Knowledge Bases
Bedrock Samples
Code examples for Bedrock APIs, agents, guardrails
AgentCore Samples
Production agent deployment, Gateway, Memory, Identity
Bedrock User Guide — Guardrails
Every filter type, PII config, ApplyGuardrail API
Bedrock User Guide — Knowledge Bases
Chunking, embeddings, vector stores
AgentCore Developer Guide
Runtime, Gateway, Policy, Memory, Observability
Well-Architected GenAI Lens
Architecture patterns tested in Domain 1
GenAI App Builder Architecture
End-to-end reference architecture
How to Choose GenAI on AWS
Service selection decision guide
GenAI Workload Assessment
Framework for evaluating implementations
Mike Chambers — AgentCore Full Walkthrough
Hands-on deploy, Gateway, Memory, Observability
Mike Chambers — AgentCore Memory Management
Short-term vs long-term memory, code walkthrough
Deploy Production AI Agents in 2 Commands
Strands Agent to AgentCore Runtime to Lambda to API Gateway
InterWorks — Bedrock Agents Walkthrough
Coffee shop agent with tool use, step-by-step console screenshots
Guardrails Deep Dive — Content Filters, PII, Streaming
Content filters, PII, streaming, prompt attacks
Guardrails Practical — PII Config, Sync vs Async
Content filters, PII config, sync vs async guardrails
PII Redaction Architecture — BDA + Guardrails
Bedrock Data Automation + Guardrails for email PII pipeline
AgentCore Policy (GA) — Cedar Policies
Cedar policies, natural language authoring, trust-but-verify
Introducing Amazon Bedrock AgentCore — Danilo Poccia
Deep walkthrough with Strands Agents code
What is Amazon Bedrock AgentCore? — Tutorials Dojo
Bedrock vs AgentCore distinction, 7 services explained
AWS Twitch — Free Live Training
Free live training with AWS experts
AWS Official Practice Set (20 questions)
20 free exam-caliber questions from AWS
CertificationPractice.com (510 questions, 6 exams)
Free — 510 questions across 6 practice exams
Tutorials Dojo Practice Exams (275+ questions)
~$15 — 275+ questions with detailed explanations
MakendranG's Handwritten Study Notes
Early Adopter badge winner's comprehensive notes
Reddit OP's AWS Service Comparison Table
Community service comparison table on Notion
Reddit OP's Service Features & Scenarios Table
Community service features and scenarios on Notion
Tutorials Dojo Study Path (Free Guide)
Free exam study path guide
Multi-Tenant RAG with JWT Isolation
Per-tenant OpenSearch isolation with KMS encryption
End-to-End RAG with CDK
IaC deployment of full RAG stack (IAM, OpenSearch, KB)
Bedrock Agents Action Groups
Lambda-backed tools for real-world tasks
Building AI Agents on AWS 2025 — Practitioner's Guide
Comparing all agent options on AWS
Multi-Agent Orchestration with Reasoning
Supervisor mode vs supervisor-with-routing mode
Multi-Agent Collaboration with Strands and Nova
Parallel agent execution and team workflows
ElastiCache Semantic Cache — 86% Cost Reduction
86% cost reduction, 88% latency improvement
Prompt Caching on Bedrock — Up to 85% Cost Reduction
Up to 85% cost reduction for repeated contexts
Effective Cost Optimization Strategies for Bedrock
Comprehensive guide to model cascading, caching, batch inference
Reinforcement Fine-Tuning — 66% Accuracy Gain
66% average accuracy gain over base models
LLM-as-a-Judge on Bedrock — 98% Cost Savings
98% cost savings vs human evaluation
Track, Allocate, and Manage GenAI Cost and Usage
Multi-team cost governance for Bedrock
AI Gateway Pattern (API Gateway + Bedrock)
Rate limiting, access control, usage tracking for GenAI APIs
Serverless GenAI Architectural Patterns
API Gateway + Lambda + Bedrock foundational patterns
Step Functions + Bedrock Orchestration
Parallel Bedrock API calls with error handling
GenAIOps — Operationalizing GenAI at Scale
Standardized repos, reusable components, automated evaluation
CloudWatch GenAI Observability
Purpose-built GenAI tracing, end-to-end across LLMs/agents/KBs
CloudWatch AppSignals for Bedrock
Native tracing, LangChain/Strands compatible
AgentCore Observability — OTEL-Based Tracing
OTEL-based tracing, debugging agent workflows
Advanced Fine-Tuning Methods on SageMaker
LoRA, QLoRA, full fine-tuning compared
Multi-LoRA Serving with vLLM on SageMaker
Dozens of adapters sharing one GPU, hot-swapping
LoRA on Trainium (HyperPod)
50% cost and 70% time reduction vs full fine-tuning
Agent Squad — Multi-Agent Orchestration
Multi-agent orchestration in Python/TypeScript
Strands Agent SOPs
Natural language agent workflows (RFC 2119 style)
AWS GenAI CI/CD Suite
Code review with Bedrock, GitHub Actions
RAG with Aurora pgvector (CDK)
Full CDK: Aurora + KB + Bedrock Agent
Evaluating LLMs Using LLM-as-a-Judge
LLM-as-a-judge sample notebooks
BDA Document Processing Samples
Bedrock Data Automation sample pipelines
Multimodal Data Processing with BDA
Guidance repo for multimodal processing pipelines
Cross-Region Inference Workshop Notebook
Jupyter notebook for CRI setup
AWS Open Source MCP Servers
MCP servers for AWS services
Strands Agents SDK Technical Deep Dive
Architecture, tool definitions, observability for Strands framework
Strands Agent SOPs — Natural Language Workflows
Natural language workflows used inside Amazon
MCP on AWS — Unlocking Model Context Protocol
AWS API MCP Server, IAM permissions, CloudTrail auditing
MCP with Bedrock Agents — Native Integration
Native MCP integration with Bedrock Agents
MCP vs Strands vs RAG vs A2A — Agents Compared
Comparison of all agent options on AWS
Automated Reasoning — 99% Verification Accuracy
Formal logic verification for mathematical correctness
Safeguard GenAI from Prompt Injections
Prompt injection defense strategies
Securing Bedrock Agents from Indirect Prompt Injections
Tag external data as 'user input' to protect agents
Intelligent Prompt Routing for Cost and Latency
Auto-routes to cheapest capable model
Evaluating RAG with Bedrock KB Evaluation
RAG-specific evaluation: retrieval relevance, answer faithfulness
Evaluate Agents with RAGAS and LLM-as-Judge
Agent evaluation: task completion, tool usage, reasoning quality
AgentCore Code Interpreter
Secure sandbox for Python execution. Supports pandas, numpy, matplotlib.
AgentCore Browser Tool
Web-based workflow execution for agents
Q Business Enterprise KB Integration
How Q Business integrates with enterprise knowledge bases
Q Developer Capabilities — Code Reviews, Docs, Tests
Code generation, completion, security scanning
MemoryDB Persistent Semantic Cache
Persistent semantic caching for GenAI workloads
Build a Proactive AI Cost Management System
Automated cost monitoring and alerting for Bedrock
EventBridge + Bedrock Event-Driven AI
Event-driven AI workflow patterns
PwC + AWS Responsible AI with Automated Reasoning
Enterprise responsible AI implementation
Automated Reasoning for Financial Services
Formal logic verification for financial calculations
Human-in-the-Loop Prompt Chaining
Step Functions Wait-for-Callback with human review
Circuit Breaker with Step Functions and DynamoDB
Resilience pattern for FM API calls
Self-Service Digital Assistant with Lex and Bedrock
Structured dialogue chatbot pattern
Tutorials Dojo Free Sampler (30 questions)
Free 30-question practice exam sampler
Converse API Guide — Carry Out a Conversation
Official guide covering Converse and ConverseStream operations — the unified, model-agnostic API for all Bedrock models
Converse API Code Examples (Python, Java, JS)
Hands-on code examples for Converse API in multiple languages — tool use, streaming, image input, guardrails integration
Converse API Hands-On Recipe (Jupyter Notebook)
Step-by-step notebook: invoke models, handle streaming, compare InvokeModel vs Converse API responses
Amazon Bedrock Deep Dive — SimpleAWS Newsletter
Comprehensive overview of Bedrock architecture: model access, inference options, agents, guardrails, and pricing tiers
Bedrock Model Selection — Solution Architect's Guide
How to choose the right FM for cost, quality, and latency — decision framework for Bedrock model families
Amazon Bedrock Complete Deep Dive — DEV Community
End-to-end Bedrock walkthrough: model access, playground, APIs, Knowledge Bases, Agents, and Guardrails
End-to-End RAG with Bedrock KB and CloudFormation
Automate full RAG deployment: S3 data source, OpenSearch vector store, embedding model, KB creation via IaC
Bedrock KB & Agents RAG Workshop (GitHub)
Hands-on workshop: build RAG pipelines, configure chunking, test retrieval, integrate with agents
Implementing RAG with Bedrock and Lambda
Serverless RAG pattern: Lambda invokes KB retrieval, augments prompt, calls FM — production-ready architecture
Advanced RAG with Terraform: Chunking, Hybrid Search, Reranking
IaC deployment of advanced RAG: 4 chunking strategies, hybrid search, Cohere reranking — retrieve 15, rerank to 5
Agentic RAG with LlamaIndex and Mistral on Bedrock
Agent-driven RAG with query decomposition, tool selection, and iterative retrieval for complex questions
re:Invent 2025 — Advanced Agentic RAG Deep Dive (AIM425)
Live coding: build intelligent event assistant with KB, Strands, AgentCore — complete dev lifecycle
re:Invent 2024 — Build Scalable RAG with Bedrock KB (AIM305)
Advanced techniques for improving RAG accuracy and cost optimization with Bedrock Knowledge Bases
Deep Dive into Vector Data Stores for Bedrock KB
Compare OpenSearch Serverless, Aurora pgvector, MongoDB Atlas, Pinecone — architecture, setup, and tradeoffs
OpenSearch Managed Cluster as Bedrock KB Vector Store
Setup guide for managed clusters: tuning flexibility, performance options, enterprise-grade vector search
Using Aurora PostgreSQL as a Bedrock KB (pgvector)
Setup pgvector extension, create schema, configure Data API, IAM role for Bedrock — ACID-compliant vector store
Vector Database Comparison — AWS Prescriptive Guidance
Side-by-side comparison of all AWS vector databases for RAG: features, scaling, cost, and use case fit
Auto-Optimize OpenSearch Vector Database
Automated HNSW tuning: ef_construction, m, ef_search, quantization — balance recall, latency, and cost
Billion-Scale Vector DB with GPU Acceleration on OpenSearch
10x faster indexing at 25% cost using NVIDIA cuVS GPU acceleration for HNSW graph construction
Build Bedrock KB with Amazon Kendra GenAI Index
Reuse Kendra's high-accuracy retrieval as Bedrock KB backend — cross-application index sharing
Optimizing Vector Search with S3 Vectors and OpenSearch
Hybrid architecture: S3 Vectors for cold storage + OpenSearch for hot queries — cost-optimized RAG
Tutorial: Building a Simple Bedrock Agent (Official)
Console walkthrough: create agent, configure instructions, add action group, test and deploy with alias
Bedrock Agents Quickstart Workshop (300-level)
Build agent for Well-Architected queries and code generation — Lambda, ECR, Python, 2-hour workshop
Introducing Strands Agents — Open Source SDK
Model-driven agent framework: build agents in a few lines of code with @tool decorators and streaming
Strands Agent Template (For the Impatient)
Minimal working Strands agent template — deploy in under 10 minutes with basic tool definitions
AgentCore CLI: Deploy an AI Agent in Four Commands
Hands-on CLI walkthrough: agentcore create, dev, deploy — Strands SDK + AgentCore Runtime in practice
Multi-Agent Collaboration — GA Announcement
Supervisor-collaborator architecture: designate supervisor agent, associate collaborators for complex workflows
Build Multi-Agent System on Bedrock — Step by Step
Complete tutorial: create supervisor + collaborator agents, configure routing, test multi-agent collaboration
re:Invent 2025 — Agents in the Enterprise: Best Practices (AIM3310)
9 rules for production agents: start small, implement OTEL, adopt multi-agent, scale with user-specific memory
Strands Labs — Experimental Agentic Development
Cutting-edge agent patterns: Python + TypeScript SDKs, experimental tools, and research-driven approaches
Build Scalable Agents with NVIDIA NeMo and AgentCore
Enterprise-scale agent deployment with NVIDIA NeMo guardrails, AgentCore Runtime, and Strands orchestration
Create Your Guardrail — All Six Policy Types
Configure content filters, denied topics, word filters, PII, contextual grounding, and Automated Reasoning
Detect Prompt Attacks with Bedrock Guardrails
Configure prompt attack filter: jailbreaks, injections, leakage detection — tag user vs system inputs
AI Safety Controls at Scale with Bedrock Guardrails
Production guardrails patterns: content moderation, PII redaction, ApplyGuardrail API, streaming integration
Build Reliable AI with Automated Reasoning — Part 1
Four-phase implementation: create policy, test with scenarios, deploy in guardrail, integrate in app
Build Responsible AI Applications with Guardrails
End-to-end responsible AI: content filters + PII + grounding checks + topic denial in one guardrail
Bedrock Guardrails Expands to Code Domain
Apply guardrails to code generation: detect insecure patterns, block credential leakage in generated code
Hacking GenAI Applications — From Theory to Practice
Red-team perspective on prompt injection attacks and defenses — practical attack/defense scenarios
Prompt Management — Construct and Store Reusable Prompts
Create prompt variants, compare side-by-side, version control, integrate with Flows and Agents
Implementing Advanced Prompt Engineering with Bedrock
Chain-of-thought, few-shot, role-play, and structured output techniques with Bedrock models
Prompt Evaluation with Prompt Management and Flows (Notebook)
Hands-on notebook: create prompt variants, build evaluation flows, compare outputs programmatically
Getting Started with Prompt Management Flows
Step-by-step: build prompt chains with Flows visual builder, integrate KB retrieval nodes
Guide to Prompt Optimization in Amazon Bedrock
Auto-rewrite prompts for better accuracy: Prompt Optimization feature reduces trial-and-error
Implement Prompt Caching on Bedrock — Cut Costs in Half
Practical implementation: cache static context, automatic vs manual checkpoints, 90% input token savings
Intelligent Prompt Routing — Understanding and Setup
Configure prompt routers: auto-route to cheapest capable model within a family, up to 30% cost savings
ElastiCache Semantic Cache — 86% Cost Reduction
Hands-on: vector embeddings for query similarity, ElastiCache as semantic cache, 86% cost and 88% latency savings
Effective Cost Optimization Strategies for Bedrock
Comprehensive guide: model cascading, prompt caching, batch inference, provisioned throughput, distillation
Optimizing Cost for FMs with Amazon Bedrock — FinOps Blog
FinOps perspective: pricing options, model selection for cost, KB optimization, automated reasoning ROI
Track, Allocate, and Manage GenAI Cost with Bedrock
Application Inference Profiles, cost allocation tags, Cost Explorer integration — per-team chargeback
Build Proactive AI Cost Management System — Part 2
Automated cost monitoring: CloudWatch alarms, rate limiting, budget alerts for Bedrock workloads
Batch Job Orchestration with Step Functions
50% savings with batch inference: S3 Map state, parallel processing, concurrency management
Bedrock Cost Optimization Techniques and Best Practices
Practical guide: on-demand vs provisioned, cross-region inference, model distillation for cost savings
VPC Endpoints (PrivateLink) for Amazon Bedrock
Private connectivity from VPC to Bedrock — never expose traffic to public internet, endpoint policies
VPC Endpoints for AgentCore Gateway
Secure agent-to-tool traffic within VPC: configure interface endpoints, restrict with endpoint policies
AWS Managed Policies for Amazon Bedrock
All predefined IAM policies: AmazonBedrockFullAccess, ReadOnly, service roles — least-privilege patterns
Security Best Practices for Fine-Tuning in Bedrock
S3 encryption, VPC configs, IAM roles, CloudTrail logging during model customization
Securing Amazon Bedrock: What Enterprises Need to Get Right
Enterprise security checklist: IAM layers, CloudTrail, CloudWatch AppSignals, compliance frameworks
AgentCore Policy — Cedar Policies for Agent Governance
Getting started: write Cedar policies via natural language or code, intercept agent-tool calls, audit trail
Understanding IAM for Managed AWS MCP Servers
IAM permission model for MCP servers: tool-level access control, CloudTrail auditing, security boundaries
RAG Evaluation and LLM-as-a-Judge on Bedrock (GA)
GA capabilities: quality, user experience, instruction compliance, safety metrics — 98% cost savings vs human eval
Evaluate Bedrock Agents with RAGAS and LLM-as-Judge
Agent evaluation: convert Bedrock messages to RAGAS format, measure tool usage, reasoning quality, task completion
Evaluate RAG with Bedrock, LlamaIndex, and RAGAS
Open-source RAG evaluation: context relevance, answer faithfulness, answer relevancy metrics with RAGAS
LLM-as-a-Judge: Automate Evaluation with Nova + Claude
Hands-on tutorial: configure LLM judge, select evaluation metrics, run automated assessment pipelines
Evaluate Model Performance Using LLM as Judge (Docs)
Official guide: set up evaluation jobs, configure judge models, interpret results, custom evaluation criteria
Orchestrate GenAI Workflows with Bedrock and Step Functions
Parallel Bedrock API calls, sequential chains, map jobs, conditions — serverless GenAI orchestration
Serverless Prompt Chaining — GitHub Repo
CDK code: sequential chains, parallel jobs, loops, conditions — Streamlit demo with meal planner example
Build GenAI Agent Workflows with Step Functions
Multi-agent Step Functions patterns: supervisor routing, parallel execution, human-in-the-loop review
Human-in-the-Loop GenAI Prompt Chaining
Wait-for-Callback pattern: task token pause, human review via SNS/SQS, resume after approval
Large-Scale Document Processing with Step Functions + Batch Inference
Map state iterates S3 objects, Bedrock batch inference, concurrency management, error handling
Autonomous Multi-Agent Workflows with Bedrock + Step Functions
Beyond chatbots: autonomous agent workflows with error recovery, parallel execution, and state management
Agentic RAG in Amazon Q Business
Q Business evolves beyond simple RAG: agent-driven retrieval, multi-step reasoning, action execution
Accuracy Evaluation Framework for Q Business
Measure Q Business response quality: relevance, accuracy, and completeness metrics with evaluation framework
April 2025: A Month of Innovation for Q Developer
Latest Q Developer features: C#/C++ support, conversation history, code reviews, security scanning
Getting Started: Q Business RAG with S3 Files
Hands-on: connect S3 documents to Q Business, configure retrieval, test conversational assistant
Q Business vs Custom RAG — AWS AI Agents Compared
Comprehensive comparison: Q Business, Bedrock KB, Strands, AgentCore, MCP — when to use which
Customize a Model with Fine-Tuning in Bedrock
Official guide: prepare training data, create fine-tuning job, configure hyperparameters, deploy custom model
Reinforcement Fine-Tuning with OpenAI-Compatible APIs
Technical walkthrough: RFT with Nova and open-weight models, 66% accuracy gains, grading functions
Model Distillation in Amazon Bedrock
Transfer knowledge from teacher to student model: automated data synthesis, cost-effective deployment
LLM Fine-Tuning on AWS — SFT, Continued Pre-Training, RLHF
Comprehensive SageMaker fine-tuning: SFT, continued pre-training, RLHF — QLoRA, LoRA, full fine-tuning compared
Import Fine-Tuned Model into Bedrock as Custom Model
SageMaker to Bedrock pipeline: fine-tune externally, import model, serve via Bedrock API
Advanced Fine-Tuning for Multi-Agent Orchestration
Amazon's internal patterns: fine-tuning models for tool selection, routing, and multi-agent coordination
Fine-tune and Deploy with SageMaker Canvas + Bedrock
No-code fine-tuning with Canvas: prepare data visually, train, deploy to Bedrock — for non-ML teams
Intelligent Document Processing at Scale with BDA
End-to-end IDP with BDA: classification, extraction, normalization, validation — reusable IaC
Programmatically Creating IDP with BDA + Strands + AgentCore
Agent-driven IDP: Strands SDK + BDA + Knowledge Base for intelligent document routing and extraction
Multimodal Power of BDA for Unstructured Data
Process documents, images, audio, video — multimodal data to structured output with BDA
IDP with Textract, Bedrock, and LangChain
OCR + LLM pipeline: Textract extracts text, Bedrock generates structured output, LangChain orchestrates
IDP Accelerator — Scalable Serverless Document Processing
Lambda + Step Functions + SQS + DynamoDB: modular IDP pipeline handling 35K-45K daily campaigns
Lessons Learned with BDA in an IDP Product
Real-world BDA experience: gotchas, workarounds, regional limits (us-west-2, us-east-1), page limits
Guidance for Deploying MCP Servers on AWS
Official architecture: containerized MCP servers on ECS/EKS/Lambda — secure cloud deployment
Run MCP Servers with AWS Lambda (GitHub)
Python library: wrap stdio MCP servers as Lambda functions, DynamoDB session management, HTTP transport
MCP Servers for ECS, EKS, and Serverless — AWS Blog
Official launch: MCP servers for container and serverless workloads with SAM CLI integration
Official AWS MCP Servers — GitHub Repository
Official AWS MCP server implementations: S3, DynamoDB, Lambda, ECS, EKS, Bedrock, and more
MCP Servers for DevOps — Complete Guide 2026
MCP patterns for DevOps: infrastructure management, CI/CD integration, monitoring — practical examples
Model Invocation Logging — CloudWatch and S3
Enable logging: JSON events to CloudWatch Logs, large I/O to S3, text/image/embedding data types
CloudWatch GenAI Observability — Official Docs
Pre-configured dashboards: invocation count, token usage, errors, latency — out-of-the-box for Bedrock
Monitoring GenAI Apps with Bedrock and CloudWatch
Hands-on: enable logging, create custom metric filters, set up alarms, build GenAI dashboards
X-Ray Integration with AgentCore
Distributed tracing for agents: propagate trace context, monitor tool calls, cross-service visibility
Configure AgentCore Observability — Official Docs
Enable OTEL instrumentation, view spans/traces, integrate with existing monitoring stack
GenAIOps on AWS: End-to-End Observability Stack — Part 3
Full observability architecture: CloudWatch + X-Ray + AppSignals + OTEL for GenAI production workloads
AgentCore Observability Cheat Sheet
Quick reference: all metrics, traces, spans, and dashboard configurations for AgentCore agents
CloudWatch Insights for Real-Time Model Invocation Analysis
Live Tail + CloudWatch Insights queries: real-time log analysis for debugging Bedrock invocations
CloudWatch GenAI Observability Samples (GitHub)
Sample dashboards, metric filters, and alarm configurations for GenAI observability on AWS
Amazon Bedrock Workshop (Comprehensive Notebooks)
Full workshop: text generation, embeddings, agents, RAG, cross-region inference — SageMaker Studio recommended
Bedrock Customization Workshop (Fine-Tuning Notebooks)
Hands-on model customization: continued pre-training, supervised fine-tuning, evaluation
Amazon Bedrock Recipes — All Hands-On Examples
Curated recipe portal: Converse API, agents, flows, guardrails, knowledge bases — runnable notebooks