Architecting Agentic AI Solutions
About This Course
As the capabilities of Generative AI mature, the industry focus is shifting from simple prompt engineering toward the development of autonomous, goal-driven Artificial Intelligence(AI) agents. These agentic systems combine Large Language Models(LLMs) with reasoning, memory, planning, and tool execution to perform complex tasks independently — from enterprise automation to customer support and domain-specific decision-making.
A recent report by Niveus Solutions projects that by 2028, 33% of enterprise software applications will incorporate Agentic AI, a sharp rise from less than 1% in 2024, indicating an accelerated demand for enterprise-ready Artificial Intelligence(AI) architecture expertise across the region [https://niveussolutions.com/emergence-of-agentic-ai-in-singapore/]. Similarly, Frontier Enterprise reports that over 70% of organizations in the Asia-Pacific region are planning to adopt Artificial Intelligence(AI) agents, signaling a strong regional shift toward agent-based Artificial Intelligence(AI) system integration [https://frontier-enterprise.com/embracing-agentic-ai-in-apacs-future-workplace/].
Despite this momentum, most practitioners lack the expertise to architect scalable, resilient, and secure Agentic AI systems. Current Large Language Models-focused training often emphasizes model usage and tuning, but not system design or deployment at scale, creating a significant talent gap.
This course is designed to bridge that gap by equipping participants with advanced skills in:
• Designing full-stack logical and physical architectures for agent-based systems
• Implementing multi-agent collaboration strategies and real-time system interoperability
• Selecting and applying appropriate Agentic AI frameworks, such as LangChain, AutoGen, OpenAI Assistants API, and Claude, for building and integrating intelligent agents into enterprise systems
What You'll Learn
• Differentiate Agentic AI from Artificial Intelligence(AI) and Robotic Process Automation(RPA) applications and identify the key agentic capabilities: Understand how Agentic AI compares to Robotic Process Automation and Generative AI, and recognize the four core agentic capabilities: reasoning, external memory, execution, and planning.
• Architect an Agentic AI solution following a development process: Learn how to select and optimize foundation models through fine-tuning or prompt engineering, implement Retrieval Augmented Generation(RAG) pipelines, manage context with vector databases, and architect AI agents for real-world data and task flows.
• Create the logical architecture of an Agentic AI solution by leveraging on key architectural styles, patterns and components: Analyze and design logical architectures that support key system qualities such as autonomy, scalability, security, fault tolerance, explainability, and adaptability; Apply proven design patterns for reflection, planning, tool use, and multi-agent collaboration to improve agent reliability, coordination, and performance.
• Select the suitable implementation frameworks for an Agentic AI solution: Gain hands-on experience with frameworks such as LangChain, AutoGen, Claude/OpenAI Assistants API, and understand their strengths, trade-offs, and use cases in enterprise contexts.
• Create the physical architecture of an Agentic AI solution by leveraging on the implementation frameworks and key architectural styles, patterns and components: Architect infrastructure for deploying agentic systems using microservices, containers, serverless models, and integrate with modern platforms like Application Programming Interface (API) gateways, service meshes, message brokers, and monitoring stacks.
Entry Requirements
Participants should have software development background with basic Artificial Intelligence(AI) knowledge.