Home Courses AWS AI
Data Science & AI ☁ AWS AI · Bedrock · SageMaker ✔ 3 Levels Available

AWS Artificial Intelligence
Full Programme

A comprehensive hands-on curriculum engineered for cloud AI practitioners and machine learning engineers. Master Amazon Bedrock foundation models, SageMaker end-to-end MLOps pipelines, computer vision and NLP cloud APIs, enterprise RAG with OpenSearch Serverless, Bedrock Agents, and production AI governance aligned with AWS Certified AI Practitioner (AIF-C01) and ML Engineer (MLA-C01/MLS-C01) standards.

🎯
Format
Theory + Lab/Assignments
🏆
Certificate
Arich Certified
Enrol Now →
Course Code
AWS-AI
Click a Level to View its Syllabus
L1
AWS-AI-AIF · Associate
6 Modules · Bedrock, SageMaker & RAG
›
L2
AWS-AI-PRO · Professional
4 Modules · Advanced MLOps & Agents
›
L3
AWS-AI-MLS · Specialty
4 Modules · Fine-Tuning & Security
›
📄 AWS AI — Associate Level (AWS-AI-AIF)
L1
AWS-AI-AIF
AWS AI & ML Foundations (AIF-C01 Aligned)
01
Generative AI & AWS Bedrock
▼
1.1 Core GenAI Principles & Foundation Models (FMs)
Transformer Architecture, Tokens, Context Windows & Temperature Parameters
Prompt Engineering Techniques on AWS (Few-Shot, CoT, System Prompts)
AWS Bedrock Guardrails: Content Filtering, Topic Blocking & PII Masking
Foundation Model Evaluation (LLM-as-a-Judge, Automated Accuracy & Robustness Metrics)
Hands-on Lab: Provision Bedrock FMs, Configure Guardrails against PII/Hallucinations, and benchmark Claude 3 vs Llama 3
02
Classical Machine Learning & Data Engineering
▼
2.1 End-to-End ML Pipeline Architecture with Amazon SageMaker
Supervised vs Unsupervised Learning: Classification, Regression, Clustering
Amazon SageMaker Data Wrangler: Visual Data Prep, Cleaning & Feature Engineering
Amazon SageMaker Feature Store: Online Low-Latency & Offline Batch Storage
SageMaker Pipelines: Automated Workflow Orchestration & Step Definitions
Hands-on Lab: Build end-to-end Automated ML Training Pipeline using Data Wrangler, Feature Store & SageMaker XGBoost
03
Core AWS Vision, NLP & Speech AI Services
▼
3.1 Specialized Pre-trained Cloud AI APIs & Serverless Integration
Amazon Rekognition: Facial Analysis, Object Detection, Moderation & OCR
Amazon Comprehend: Entity Recognition, Keyphrase Extraction & Sentiment Analysis
Amazon Transcribe & Polly: Real-Time Audio-to-Text & Neural Text-to-Speech
Amazon Textract & Translate: Intelligent Document Processing & Neural Translation
Hands-on Lab: Build Automated Document & Audio Analytics Workflow using AWS Lambda, Rekognition, Comprehend, and Transcribe
04
RAG Architectures & Vector Databases
▼
4.1 Advanced Knowledge Ingestion & Retrieval Architectures
Retrieval-Augmented Generation (RAG) Patterns with Foundation Models
Amazon OpenSearch Serverless Vector Engine & Vector Indexes
Amazon Bedrock Knowledge Bases: Data Sources (S3), Chunking & Parsing Strategies
Titan Embeddings, Cohere Rerank & Hybrid Lexical-Vector Retrieval
Hands-on Lab: Construct Enterprise RAG Pipeline using Bedrock Knowledge Bases, OpenSearch Vector Search, and Re-ranking
05
Autonomous Bedrock Agents & Tool Integration
▼
5.1 Autonomous Cloud Agent Architectures & Orchestration
Amazon Bedrock Agents: Foundation Models + Action Groups + Knowledge Bases
OpenAPI 3.0 Tool Definitions & AWS Lambda Tool Ingestion
ReAct Cognitive Loops, Plan Breakdown & Autonomous Multi-Step Tool Execution
Session State Persistence & Memory Management in Bedrock
Hands-on Lab: Build an Autonomous Customer Support Bedrock Agent with DynamoDB & Lambda Action Groups
06
Enterprise GenAI Security, Fine-Tuning & MLOps Governance
▼
6.1 Production GenAI Governance & Customization Lifecycles
Amazon Bedrock Custom Models: Fine-Tuning vs Continued Pre-Training with LoRA / PEFT
AWS IAM AI Policies, VPC Endpoints & KMS Key Encryption for Foundation Models
Amazon CloudWatch GenAI Metrics, Telemetry & Token Cost Optimization
Amazon SageMaker Model Cards, Clarify Bias Checks & Governance Workflows
Capstone Project: Deploy Fine-Tuned Custom Model with Real-Time Endpoint, Model Cards, and IAM Security Boundaries
📄 AWS AI — Professional Level (AWS-AI-PRO)
L2
AWS-AI-PRO
Advanced MLOps, Distributed Training & Cloud Agents (MLA-C01 Aligned)
01
Advanced SageMaker Feature Engineering & Distributed Training
▼
SageMaker Processing Jobs for Large-Scale Data Preprocessing with PySpark
Distributed Training Architectures: Data Parallelism & Model Parallelism on AWS Trainium / GPU clusters
SageMaker Hyperparameter Optimization (HPO) with Bayesian Search & Early Stopping
Model Compilation & Quantization using AWS Inferentia / AWS Neuron SDK
Hands-on Lab: Distributed Model Training Pipeline on Multi-Node AWS Trainium / GPU Instances
02
Enterprise MLOps & Automated CI/CD Pipelines
▼
SageMaker Model Registry: Versioning, Approval Status & Metadata Tracking
Automated CI/CD Workflows with AWS CodePipeline, CodeBuild & GitHub Actions
Deployment Strategies: Real-Time Endpoints, Serverless Inference, Asynchronous Inference & A/B Shadow Testing
SageMaker Model Monitor: Real-Time Data Drift, Concept Drift & Model Quality Alerts
Hands-on Lab: Complete Production SageMaker CI/CD Pipeline with Automated Model Drift Retraining
03
Multi-Modal GenAI & Advanced Vector Search
▼
Multi-Modal Embeddings with Amazon Titan Multimodal & Claude 3 Vision APIs
Amazon OpenSearch Serverless k-NN Index Optimization & Hierarchical Navigable Small World (HNSW) graphs
Hybrid Search: Combining BM25 Lexical Matching with Semantic Vector Similarity
Context Window Compaction, Metadata Filtering & Query Transformation Pipelines
Hands-on Lab: Build a Multi-Modal Visual Search Engine combining Text & Image Embeddings on AWS OpenSearch
04
Production Bedrock Agent Orchestration & Enterprise Integrations
▼
Multi-Agent Collaboration & Supervisor Patterns on Amazon Bedrock
Integrating Bedrock Agents with Enterprise Systems: Salesforce, Snowflake, Jira & SAP via Action Groups
Asynchronous Human-in-the-Loop (HITL) Approval Queues with Amazon SQS & Step Functions
Agent Telemetry, Prompt Caching & Latency Optimization on AWS CloudWatch
Hands-on Lab: Construct an Enterprise IT Service Desk Multi-Action Agent with Asynchronous HITL Workflows
📄 AWS AI — Specialty Level (AWS-AI-MLS)
L3
AWS-AI-MLS
Enterprise AI Architecture, Zero-Trust Security & Custom Alignment (MLS-C01 Aligned)
01
High-Performance GenAI Architecture & Custom Model Adaptation
▼
Continued Pre-training vs Instruction Fine-Tuning on SageMaker JumpStart & Bedrock
Parameter-Efficient Fine-Tuning (PEFT), LoRA & QLoRA on Custom AWS Infrastructure
Direct Preference Optimization (DPO) & RLHF on Custom Foundation Models
Model Evaluation Frameworks: Automated Benchmarking, Perplexity & Task-Specific Accuracy
Hands-on Lab: Enterprise Domain Fine-Tuning & DPO Alignment on AWS SageMaker with Custom Evaluation Benchmarks
02
Enterprise GenAI Security, Zero-Trust & Compliance
▼
Zero-Trust AI Architecture: AWS PrivateLink, VPC Endpoints & Isolated Foundation Model Ingestion
Customer-Managed Encryption Keys (AWS KMS) for Bedrock FMs, Knowledge Bases & Embeddings
Advanced Bedrock Guardrails: Automated PII Masking, Hallucination Prevention & Regex Safety Rules
Compliance & Governance: AWS Audit Manager, SageMaker Clarify Bias Audits & Model Cards
Hands-on Lab: Build a Zero-Trust GenAI Security Architecture with KMS Encryption, VPC Endpoints & Automated PII Masking
03
Advanced Cognitive Graph-RAG & Agentic Memory on AWS
▼
Graph-RAG Integration: Connecting Amazon Neptune Graph Database with Bedrock Knowledge Bases
Combining Knowledge Graphs with Vector Embeddings for Complex Entity Reasoning
Long-Term Agentic Memory Architecture using Amazon DynamoDB & Amazon ElastiCache Redis
Tree-of-Thought (ToT) & Self-Correcting Reflection Loops for Autonomous Agents
Hands-on Lab: Build a Graph-RAG Financial Intelligence Pipeline on AWS Neptune + Bedrock with Multi-Hop Reasoning
04
Cloud AI FinOps, Observability & Enterprise Architecture Capstone
▼
AWS Bedrock Provisioned Throughput Planning, Capacity Reservation & Token Cost Forecasting
CloudWatch GenAI Dashboards, OpenTelemetry Tracing & Latency SLAs
Multi-Account Governance with AWS Organizations & Service Control Policies (SCPs) for AI
Capstone Project: End-to-End Enterprise Generative AI Platform on AWS with Bedrock, SageMaker, OpenSearch & Full MLOps Telemetry