Certified AI Engineering Professional

Topic 1.1: Artificial Intelligence Foundations
 

Artificial Intelligence fundamentals
Machine Learning fundamentals
Deep Learning fundamentals
Generative AI fundamentals
Neural Networks overview
Foundation Models

Topic 1.2: Role of an AI Engineer Subtopics 

Responsibilities of an AI Engineer
AI Engineer vs ML Engineer
AI Engineer vs Data Scientist
Skills required for modern AI Engineering Topic

Topic 1.3: Modern AI Engineering Ecosyste

Large Language Models 
RAG and AI Agents 
AI Engineering lifecycle 
Modern AI application architecture
AI engineering tools and platforms
From prototype to production AI

Topic 2.1: Transformer Fundamentals

Why transformers changed modern AI
Encoder, decoder and transformer data flow
Positional information and token order

Topic 2.2: Attention & Self-Attention

Attention, query, key and value concepts
Self-attention and multi-head attention
Causal attention for text generation


Topic 2.3: Tokens & Context Windows

Tokens and tokenization
Context windows and long-context limits
Token usage, latency and cost impact

Topic 2.4: Large Language Models

Pretraining, instruction tuning and inference
Open-weight vs proprietary models
Reasoning and multimodal models

Topic 2.5: Model Selection

Model selection by quality, latency and cost
Benchmarking models for a real workload
When to use small vs large models

Topic 3.1: Working with LLM APIs

LLM API request-response lifecycle
Authentication, streaming and rate limits
Retries, errors and resilient API integration

Topic 3.2: Model Parameters

Temperature, top-p and output length
Parameter settings for different workloads

Topic 3.3: System & User Prompts

System prompts and user prompts
Instruction hierarchy and prompt structure

Topic 3.4: Structured Outputs

JSON and schema-based outputs
Validation and recovery from malformed output

Topic 3.5: Function & Tool Calling

Function schemas and tool definitions
Tool selection, execution and result handling
Safe error handling for tool calls

Topic 3.6: Context Management

Conversation history and context trimming
Memory vs temporary context

Topic 3.7: Routing & Fallbacks

Model routing and fallback patterns
Building a reliable multi-model application

Topic 4.1: Prompt Engineering Fundamentals

Anatomy of a strong production prompt
Instructions, constraints and success criteria
Prompt testing and iterative improvement

Topic 4.2: Zero-Shot & Few-Shot Prompting

Zero-shot prompting
Few-shot prompting with representative examples
Choosing examples and avoiding bias

Topic 4.3: Prompt Templates & Chaining

Reusable prompt templates
Prompt chaining for complex tasks
Conditional and multi-step prompt workflows

Topic 4.4: Context Engineering

Selecting and ordering useful context
Managing conflicting and untrusted context

Topic 4.5: Fine-Tuning Fundamentals

What fine-tuning changes
When fine-tuning is appropriate
Training data and evaluation considerations

Topic 4.6: LoRA & PEFT

LoRA and parameter-efficient fine-tuning
Prompting vs RAG vs fine-tuning decision framework

Topic 5.1: Understanding Embeddings

Embeddings and semantic meaning
Vector dimensions and semantic similarity
Common embedding use cases

Topic 5.2: Embedding Models

Choosing an embedding model
Embedding versioning and re-indexing implications

Topic 5.3: Semantic & Similarity Search

Semantic search workflow
Cosine similarity and top-k retrieval
Similarity thresholds and relevance

Topic 5.4: Vector Databases

How vector databases store and index data
Collections, namespaces and indexes
Selecting a vector database

Topic 5.5: Metadata Filtering

Metadata design and filtering
Tenant and access-aware retrieval

Topic 5.6: Dense, Sparse & Hybrid Search

Dense vs sparse retrieval
Hybrid search and rank fusion

Topic 6.1: RAG Fundamentals & Architecture

Why RAG is needed
RAG components and end-to-end request flow
Offline indexing vs online retrieval
Common RAG failure points

Topic 6.2: Document Loading & Parsing

Loading PDF, DOCX, HTML and structured content
Preserving headings, tables and metadata
Handling poor-quality or malformed documents

Topic 6.3: Document Chunking

Fixed, recursive and structure-aware chunking
Chunk size, overlap and information continuity
Choosing chunking by document type

Topic 6.4: Embedding & Indexing

Embedding documents and storing metadata
Index creation, updates and re-indexing

Topic 6.5: Vector Retrieval

Query embedding and top-k retrieval
Metadata-aware retrieval and thresholds
Diagnosing irrelevant retrieval

Topic 6.6: Context Augmentation

Formatting and ordering retrieved evidence
Reducing duplicated or conflicting context

Topic 6.7: Source Citations

Tracking source provenance
Generating traceable cited answers

Topic 6.8: End-to-End RAG Build

Build the ingestion-to-answer pipeline
Evaluate a working RAG application

Topic 7.1: Advanced Chunking

Semantic and hierarchical chunking
Parent-child retrieval patterns

Topic 7.2: Query Rewriting

Query expansion and rewriting
Multi-query and conversation-aware retrieval

Topic 7.3: Hybrid Retrieval

Combining semantic and keyword retrieval
Rank fusion and weighting strategies

Topic 7.4: Metadata-Based Retrieval

Business, tenant and access filters
Time and category-aware retrieval

Topic 7.5: Reranking

Why reranking improves retrieval quality
Cross-encoder and model-based reranking

Topic 7.6: Context Optimization

Removing redundancy and compressing context
Selecting the best evidence for generation

Topic 7.7: Conversational & Multi-Source RAG

RAG with conversation history
Retrieving from multiple repositories

Topic 7.8: Corrective, Adaptive & GraphRAG

Corrective and adaptive RAG patterns
Graph RAG and knowledge graph overview

Topic 8.1: AI Agent Fundamentals

AI agents vs chatbots and workflows
Goals, observations, actions and feedback loops
Enterprise agent use cases

Topic 8.2: Agent Architecture

Model, tools, memory, state and controller
Planner-executor and orchestration patterns

Topic 8.3: Agent Planning & Decision Making

Task decomposition and action selection
Stopping conditions and loop prevention

Topic 8.4: Tool & Function Calling

Designing safe agent tools
Validation, permissions and tool failures

Topic 8.5: Agent Memory & State

Short-term and long-term memory
Structured state and memory retrieval

Topic 8.6: Agent Workflows

Sequential, conditional and parallel workflows
Retries, timeouts and compensation patterns

Topic 8.7: Human-in-the-Loop

Approval gates for sensitive actions
Escalation and auditability

Topic 8.8: Multi-Agent Systems

Specialist and supervisor agent patterns
Multi-agent coordination risks and trade-offs

Topic 8.9: Build a Tool-Using Agent

Implement a bounded tool-using agent
Test agent success, failure and safety scenarios
 

Topic 9.1: Traditional vs Agentic RAG

Traditional RAG vs agentic RAG
When agent-controlled retrieval adds value

Topic 9.2: Agent-Controlled Retrieval

Agent decides whether and where to retrieve
Iterative retrieval and stopping criteria

Topic 9.3: Dynamic Knowledge Retrieval

Retrieval from APIs, databases and live systems
Freshness, caching and failure handling

Topic 9.4: Tool-Based Retrieval

Exposing search and retrieval as agent tools
Validating tool-retrieved evidence

Topic 9.5: MCP Fundamentals

What Model Context Protocol standardizes
MCP hosts, clients and servers
MCP tools, resources and prompts

Topic 9.6: MCP Integration

Connecting agents to MCP servers
Discovering and using MCP capabilities
Designing a practical multi-MCP architecture

Topic 9.7: MCP Production Architecture & Security

Stateless MCP architecture and scalable request handling
MCP authorization, identity and least-privilege access
MCP Tasks and long-running operations

Topic 10.1: LLM Evaluation

Task-specific evaluation criteria
Evaluation datasets and regression testing
Human, rubric and semantic evaluation

Topic 10.2: RAG Evaluation

Retrieval relevance and precision
Faithfulness, groundedness and answer quality

Topic 10.3: Agent Evaluation

Task success and tool-selection accuracy
Action correctness, loops and recovery

Topic 10.4: Hallucination & Groundedness

Types of hallucination
Grounding responses in evidence

Topic 10.5: LLM-as-a-Judge

Model-based evaluation and rubrics
Bias and calibration of model judges

Topic 10.6: AI Guardrails

Input, output and policy guardrails
Layered guardrail architecture

Topic 10.7: Prompt Injection Protection

Direct and indirect prompt injection
Protecting RAG and tool-enabled applications

Topic 10.8: Data Privacy, Responsible AI & Risk Management

Sensitive data, secrets and tenant isolation
Bias, transparency, human oversight and AI risk-management checkpoints

Topic 10.9: AI Application Security Risks

OWASP Top 10 for LLM Applications 2026 overview
Excessive agency, improper output handling and unbounded consumption
Vector, embedding, hidden-context and supply-chain security risks

Topic 11.1: LLMOps Fundamentals

LLMOps lifecycle and production concerns
Development, staging and production environments

Topic 11.2: Prompt & Model Versioning

Versioning prompts, models and configurations
Rollback and change-control strategy

Topic 11.3: Logging & Tracing

Request logging and correlation IDs
Tracing RAG, agent and tool execution

Topic 11.4: AI Observability

Quality, retrieval and agent monitoring
Continuous evaluation, AI health monitoring and alerting

Topic 11.5: Token, Latency & Cost Monitoring

Token and latency metrics
Cost per request and budget monitoring

Topic 11.6: Model Serving

Hosted vs self-hosted model serving
Concurrency, batching and caching

Topic 11.7: Docker & Cloud Deployment

Containerizing an AI application
Secrets, configuration and cloud deployment

Topic 11.8: CI/CD for AI Applications

Automated evaluation as a quality gate
Progressive deployment and rollback

Topic 11.9: Performance, Cost & Reliability

Model routing and token optimization
Retries, timeouts, fallbacks and circuit breakers

Topic 12.1: Project Architecture

Define the enterprise AI use case
Design the end-to-end architecture

Topic 12.2: RAG Implementation

Build ingestion, retrieval and cited answers
Evaluate RAG quality

Topic 12.3: Agent & Tool Integration

Connect external tools and APIs
Add bounded agent actions and approvals

Topic 12.4: MCP Integration

Connect the application to an MCP server
Use MCP tools or resources in the workflow

Topic 12.5: Evaluation & Guardrails

Create evaluation scenarios
Add guardrails and prompt-injection tests

Topic 12.6: Deployment & Final Demo

Containerize and deploy the application
Demonstrate production readiness, observability and safe operation