AI Agents & Agentic Workflow Development Certification

Module 1: Introduction to AI Agents

• Understanding AI agents
• Agents versus traditional applications
• Agent capabilities
• Autonomous and semi-autonomous systems
• Common agent use cases

Module 2: Agentic AI Fundamentals

• Goals and tasks
• Reasoning and actions
• Environment interaction
• Feedback loops
• Agent lifecycle

Module 3: Large Language Models for Agents

• LLM capabilities
• Context and instructions
• Model inputs and outputs
• Model limitations
• Selecting models for agent tasks

Module 4: Agent Architecture

• Core architecture components
• Model layer
• Tool layer
• Memory layer
• Orchestration layer

Module 5: Designing Agent Instructions

• System instructions
• Defining agent roles
• Setting objectives
• Establishing constraints
• Designing reliable instructions

Module 6: Advanced Prompt Design

• Context construction
• Task decomposition
• Few-shot examples
• Structured prompting
• Prompt optimization

Module 7: Structured Outputs

• Output schemas
• JSON-based responses
• Data validation
• Parsing model outputs
• Handling malformed responses

Module 8: Tool-Enabled Agents

• Understanding tool use
• Defining available tools
• Tool selection
• Passing arguments
• Processing tool results

Module 9: Function and API Integration

• API fundamentals
• Function definitions
• Request and response handling
• Authentication concepts
• Integrating external services

Module 10: Agent Reasoning Workflows

• Breaking down complex tasks
• Planning actions
• Executing steps
• Reviewing results
• Handling incomplete information

Module 11: Agentic Workflow Design

• Sequential workflows
• Conditional workflows
• Parallel tasks
• Routing patterns
• Workflow state management

Module 12: Workflow Orchestration

• Coordinating workflow steps
• Managing dependencies
• Controlling execution
• Passing state between steps
• Handling workflow completion

Module 13: Agent Memory Fundamentals

• Short-term context
• Persistent information
• Conversation state
• Memory retrieval
• Memory management strategies

Module 14: Context Management

• Context windows
• Selecting relevant information
• Context compression
• Managing long interactions
• Preventing context overload

Module 15: Retrieval-Augmented Agents

• Retrieval concepts
• Knowledge sources
• Document retrieval
• Providing retrieved context
• Grounding agent responses

Module 16: Embeddings and Vector Search

• Embedding concepts
• Vector representations
• Similarity search
• Vector databases
• Retrieval pipelines

Module 17: Building Knowledge-Based Agents

• Knowledge ingestion
• Document processing
• Retrieval workflows
• Answer generation
• Source-grounded responses

Module 18: Agent State Management

• Understanding state
• Session state
• Workflow state
• Updating state
• Recovering state

Module 19: Planning and Task Decomposition

• Understanding complex objectives
• Breaking tasks into subtasks
• Sequencing actions
• Managing dependencies
• Replanning when needed

Module 20: Human-in-the-Loop Workflows

• Approval checkpoints
• User confirmation
• Escalation patterns
• Manual intervention
• Balancing automation and control

Module 21: Multi-Agent Systems

• Multi-agent concepts
• Specialized agents
• Agent responsibilities
• Agent communication
• Coordinating agent activities

Module 22: Agent Routing and Delegation

• Request classification
• Selecting specialized agents
• Delegating tasks
• Combining outputs
• Managing routing failures

Module 23: Event-Driven Agent Workflows

• Events and triggers
• Scheduled workflows
• External system events
• Asynchronous processing
• Workflow continuation

Module 24: Business Process Automation

• Identifying automation opportunities
• Mapping business processes
• Integrating AI decisions
• Automating repetitive tasks
• Designing exception handling

Module 25: Error Handling and Recovery

• Identifying failure points
• Retry strategies
• Timeouts
• Fallback mechanisms
• Graceful failure handling

Module 26: Agent Reliability

• Reducing unpredictable behaviour
• Input validation
• Output validation
• Limiting agent actions
• Designing reliable workflows

Module 27: Security for AI Agents

• Access control
• Protecting credentials
• Tool permissions
• Prompt injection awareness
• Secure agent design

Module 28: Guardrails and Controls

• Defining operational boundaries
• Input controls
• Output controls
• Tool restrictions
• Escalation mechanisms

Module 29: Testing AI Agents

• Unit-level testing
• Workflow testing
• Scenario testing
• Edge cases
• Regression testing

Module 30: Agent Evaluation

• Defining evaluation criteria
• Task completion
• Response quality
• Tool-use accuracy
• Evaluating workflow outcomes

Module 31: Observability and Monitoring

• Execution traces
• Logging
• Tool-call monitoring
• Error tracking
• Performance monitoring

Module 32: Cost and Performance Optimization

• Model selection
• Token efficiency
• Reducing unnecessary calls
• Latency optimization
• Balancing quality and cost

Module 33: Deploying Agentic Applications

• Development environments
• Configuration management
• Deployment architecture
• Environment variables
• Production considerations

Module 34: Scaling Agentic Systems

• Concurrent workloads
• Queue-based processing
• Rate limits
• Resource management
• Designing scalable architectures

Module 35: Building an End-to-End AI Agent

• Defining the use case
• Designing architecture
• Connecting tools
• Implementing memory
• Testing agent behaviour

Module 36: Building an Agentic Business Workflow

• Mapping workflow requirements
• Designing agent interactions
• Integrating external systems
• Adding human checkpoints
• Evaluating workflow performance

Module 37: Production Agent Patterns

• Router patterns
• Supervisor patterns
• Specialist agent patterns
• Tool-using agent patterns
• Workflow-based architectures

Module 38: Debugging Agentic Systems

• Diagnosing incorrect outputs
• Debugging tool calls
• Reviewing workflow state
• Identifying instruction conflicts
• Improving reliability

Module 39: Responsible Agent Development

• Human oversight
• Transparency
• Data handling
• Managing automation risks
• Responsible deployment practices

Module 40: Designing Production-Ready Agentic Solutions

• Requirements analysis
• Architecture selection
• Reliability planning
• Deployment considerations
• Continuous improvement