What you'll learn
- Differentiate prompts, fixed workflows and autonomous agents and decide when each is appropriate
- Explain and trace the perceive–plan–act loop that underpins every agent
- Design and describe tool schemas and use function calling to let agents act on the world
- Manage agent memory and engineer multi-step planning with reflection and recovery
- Integrate private or current knowledge into an agent through retrieval-augmented generation (RAG)
- Build a working no-code automation that includes AI decision steps, branching and error handling
- Compare single-agent and multi-agent topologies and outline what orchestration frameworks provide
- Evaluate agent reliability with test sets and metrics, and implement guardrails & human-in-the-loop checkpoints
- Optimise an agent for cost, latency or quality and justify trade-offs for a given use-case
- Deliver a documented, evaluated and guard-railed agent/automation for a real workflow
Curriculum
Module 1: From Prompts to Agents3 topics
- Spectrum of Autonomy: Prompt, Workflow, Agent
- Agent Anatomy: The Perceive–Plan–Act Loop
- Quiz: Module 1: From Prompts to Agents
Module 2: Tool Use & Function Calling4 topics
- Why Agents Need Tools
- Designing Tool Schemas & Handling Function Calls
- Safety in Tool Design
- Quiz: Module 2: Tool Use & Function Calling
Module 3: Memory, State & Multi-Step Reasoning4 topics
- Working Context vs Long-Term Memory
- Planning & Interleaving Reason and Action
- Reflection & Failure Recovery
- Quiz: Module 3: Memory, State & Multi-Step Reasoning
Module 4: Retrieval-Augmented Agents3 topics
- Retrieval as a Tool in the Loop
- Grounding, Citation & Handling Knowledge Gaps
- Quiz: Module 4: Retrieval-Augmented Agents
Module 5: No-Code Workflows & Automation4 topics
- Anatomy of a No-Code Workflow
- Embedding AI Decision Steps
- Branching, Error Handling & Human Checkpoints
- Quiz: Module 5: No-Code Workflows & Automation
Module 6: Single vs Multi-Agent Systems & Orchestration3 topics
- Choosing the Right Topology
- Orchestration Frameworks & Coordination Risks
- Quiz: Module 6: Single vs Multi-Agent Systems & Orchestration
Module 7: Evaluation, Guardrails & Safety3 topics
- Evaluating Non-Deterministic Agents
- Stakes Ladder, Guardrails & Human-in-the-Loop
- Quiz: Module 7: Evaluation, Guardrails & Safety
Module 8: Optimisation & Capstone Integration3 topics
- Cost–Latency–Quality Trade-Offs & Optimisation Levers
- Capstone Design Clinic
- Quiz: Module 8: Optimisation & Capstone Integration
About this course
This course moves learners from simply prompting a language model to designing, building, evaluating and safely deploying autonomous AI agents and no-code automations that plan, use tools and act unattended in real-world workflows.
Prerequisites
- •Fluency writing effective chat-LLM prompts (role/task/context, few-shot examples)
- •Basic familiarity with JSON and reading simple code snippets