Search results
Agentic AI in Practice - Spring Edition
What is the next step beyond basic Generative AI? UN Agentic AI in Practice is designed for United Nations (UN) personnel seeking to explore how AI agents can go beyond generating content to performing tasks, using tools, and coordinating complex workflows.
Through a combination of self-paced learning, instructor-led webinars, and practical activities tailored to the UN context, participants will learn how to identify opportunities for Agentic AI, design no-code AI workflows, and apply these technologies responsibly to enhance productivity, knowledge management, reporting, and decision-making.
Artificial Intelligence (AI) is evolving rapidly from systems that generate information to systems that can support and execute work. While the first generation of Generative AI focused primarily on generating content in response to prompts, recent developments have introduced AI agents capable of using tools, following workflows, and performing multi-step tasks. Building on these capabilities, Agentic AI enables multiple specialised agents to collaborate towards more complex goals.
This evolution creates significant opportunities for the United Nations (UN) system to enhance productivity, knowledge management, reporting, communications, and decision-making. At the same time, as AI systems become increasingly autonomous, questions of accountability, transparency, safety, and human oversight become increasingly important.
To address these opportunities and challenges, the UN Agentic AI in Practice course is designed to equip UN personnel with the knowledge and practical skills to understand and apply Agentic AI in real UN work contexts. Participants will explore how agentic systems work, identify suitable use cases within their own functions, and learn to design and evaluate simple agentic workflows using no-code tools.
The course places particular emphasis on responsible AI adoption, enabling participants to explore the potential of Agentic AI while ensuring that AI systems operate within appropriate boundaries and remain accountable to human judgement.
Upon successful completion of the programme, participants will be able to:
- Describe the principles and elements that define the Agent Paradigm and explain how agents improve from standard generative AI tools.
- Identify and prioritise tasks in daily work that could benefit from an agentic approach and articulate the potential value for their unit.
- Use no-code agentic platforms to design and direct simple agent workflows applied to real UN work contexts.
- Evaluate agentic systems critically, identify risks related to autonomy, tool use, and apply principles of safe deployment such as bounded autonomy, human oversight, and governance requirements.
This course offers a unique blended learning programme on Agentic AI in the UN context. Participants will be guided through a curriculum developed by UN AI experts, combining self-paced learning modules with instructor-led webinars. Practical case studies and hands-on activities are key features of the programme, ensuring that participants not only acquire knowledge and skills but also apply them to real-world UN work scenarios.
For detailed information on the course content and methodology, please refer to the SYLLABUS.
The programme includes:
- Four thematic self-paced modules.
- Four instructor-led webinars.
- A final assignment.
The course emphasises the use of accessible no-code AI tools, enabling participants to design and manage agentic workflows through natural language instructions rather than traditional programming.
The four 90-minute instructor-led webinars will be conducted via Zoom from 15:00 to 16:30 (Rome time).
Participants will need a computer or mobile device, a reliable internet connection, and a headset with a microphone to participate in the live webinars.
The self-paced modules and discussion forums will be hosted on the UN Learn platform, where participants can access learning materials and engage in discussions throughout the programme.
The self-paced learning component covers four thematic modules:
- From Chats to Tasks: The Evolution of Generative AI
- Decomposing an AI Agent: How Agents Work
- Understanding Workflows in Agentic AI
- Governance, Risks, and Responsible Use of Agentic AI
The instructor-led webinars will focus on:
- Understanding Agentic AI and Its Core Components
- Exploring Agentic Platforms in Practice
- Designing an Agentic AI Workflow End to End
- Testing and Evaluating Agentic AI Workflows
The programme concludes with a final assignment, allowing participants to apply their learning by documenting an agentic workflow and critically reflecting on its design, potential applications, and governance requirements.
This course is designed for UN personnel across all levels, at headquarters and in field locations, who are interested in exploring the practical applications of Agentic AI in their work.
It is particularly relevant for:
- UN personnel with a working familiarity with Generative AI concepts and tools who wish to advance their skills in designing and using AI agents.
- Professionals involved in communications, reporting, knowledge management, programme management, learning, evaluation, and other functions where Agentic AI could support operational workflows.
- Project managers, team leaders, and decision-makers seeking to understand how to leverage and oversee AI-powered workflows responsibly.
No coding experience or specific technical background is required. However, participants are expected to have a basic familiarity with Generative AI tools, such as chatbots and prompting.
The course fee is $1,500
