Applied Gen AI and Agentic Systems
Get the Brochure
To receive a brochure with the full course content, please complete the following form with your correct details
Program Professional Framework
Applied Gen AI and Agentic Systems
Program Overview
AI agents go beyond answering questions: they plan steps, call tools and APIs, read and write data and complete multi-step tasks with limited supervision. This course explains how agentic systems are designed and governed, covering agent architectures, tool use and function calling, memory and retrieval, multi-agent orchestration, and protocols such as the Model Context Protocol for connecting agents to enterprise systems. Participants build agents for tasks like procurement request handling or report preparation, then test them for reliability, security and cost, applying guardrails, human approval steps and audit logging suitable for regulated environments.
Learning Outcomes
- ✓Explain the components of an agentic system including planner, tools, memory and control loop.
- ✓Build a single agent that uses function calling to retrieve data and complete a defined business task.
- ✓Design a multi-agent workflow with clear roles, hand-offs and human approval checkpoints.
- ✓Evaluate agent performance using task success rate, error analysis, latency and cost per task.
- ✓Apply guardrails against prompt injection, excessive permissions and unintended actions.
- ✓Define logging and oversight requirements for deploying agents in regulated processes.
Target Audience
- AI engineers and software developers
- Solution architects and automation specialists
- Data scientists moving into generative AI applications
- Digital transformation leads evaluating agent use cases
Detailed Program Modules
01From Generative AI to Agents
- Chatbots, copilots and autonomous agents
- Where agents add value and where they fail
- Selecting suitable agent use cases
02Agent Architecture
- Reasoning and planning patterns
- Tool use and function calling
- Short-term and long-term memory
03Connecting Agents to Enterprise Systems
- APIs, databases and document stores
- Model Context Protocol and tool servers
- Permission models and least privilege
04Multi-Agent Orchestration
- Role-based agent teams
- Hand-offs, routing and supervision
- Human-in-the-loop approval steps
05Testing and Evaluation
- Task success and failure analysis
- Benchmarking latency and cost
- Regression testing for agent behaviour
06Security and Governance
- Prompt injection and data exfiltration risks
- Guardrails, sandboxing and audit logs
- Operating agents in regulated environments
Training Methodology
Applied Project & Assessment
Applied Project
Participants design and build an agent or multi-agent workflow for a real process in their organization, document its architecture and permissions, test it against a set of scenarios and present results with a governance and deployment plan.
Assessment
Short diagnostic at the start, practical tasks linked to each learning outcome, and evaluation of the applied project against published criteria.
University Certificate & International Accreditation
Professional Certificate accredited by Universiti Teknologi Malaysia (UTM) in Applied Gen AI and Agentic Systems
International accreditation: American Association of Innovation (AAI)