AI Workshop
LIVE INSTRUCTOR-LED
Build your Personal AI Agents
Create personal AI agents that help you organise recurring work, apply your preferred instructions and handle repeatable tasks more consistently.
Turn your recurring tasks and preferred ways of working into reusable personal AI assistants.
1 day Team Workshop Activities and Re-usable content Certificate of participation
Workshop Overview
Overview
Many people use AI as a blank chat window each time they need help. Personal AI agents take the next step by packaging recurring context, instructions and workflow expectations into reusable assistants.
Build your Personal AI Agents is a hands-on workshop that helps participants identify repeatable tasks, define an agent's job, design clear instructions, provide useful reference material and test performance across realistic scenarios.
The workshop is tool-flexible and focuses on transferable agent design principles that can be applied in supported enterprise AI platforms.
Key Outcomes
By the end of this workshop, you'll be able to:
Identify recurring tasks that are suitable for a personal AI agent.
Distinguish an agent use case from a one-off prompt.
Define an agent's purpose, users, scope, inputs and expected outputs.
Write strong system-style instructions and behavioural guardrails.
Structure reusable context and reference knowledge.
Design repeatable interaction patterns for common tasks.
Test an agent against normal, edge and failure scenarios.
Refine instructions based on observed performance.
Recognise privacy, security and data-handling considerations.
Build a small personal portfolio of agent ideas and one working prototype.
Workshop Agenda
Explore the workshop
1 From Prompts to Personal Agents
Understand what changes when instructions and context become reusable.
2 Find the Right Task
Identify repetitive, high-friction work that benefits from consistent AI support.
3 Agent Design Canvas
Define purpose, scope, inputs, outputs, boundaries and success criteria.
4 Write Agent Instructions
Create clear role, behaviour, process and quality expectations.
5 Add Context and Knowledge
Decide what information the agent needs and how it should use it.
6 Design the Interaction
Shape how users provide inputs and how the agent returns useful outputs.
7 Test and Refine
Challenge the agent with realistic cases and improve weak behaviour.
8 Responsible Personal Agents
Consider sensitive information, permissions, accuracy and human review.
9 Build Your Agent Portfolio
Capture additional agent opportunities and next steps.
Who It's For
Built for people doing the work
Business professionals
Managers and leaders
Consultants
Project and delivery professionals
Product managers
Analysts
Operations professionals
AI champions
Anyone who repeatedly uses AI for similar tasks
Hands-on Activities
Learn by doing
Practical exercises you'll complete during the workshop.
Agent Design Canvas
Instruction-writing exercise
Reference knowledge design
Interaction template creation
Agent prototype build
Agent test challenge
Refinement cycle
Personal agent portfolio map
You'll Leave With
Practical assets you can reuse
Personal AI Agent Prototype
Agent Design Canvas
Reusable Instruction Template
Knowledge and Context Checklist
Interaction Template
Agent Test Script
Responsible Use Checklist
Personal Agent Opportunity Portfolio
Certification
Certificate of completion
Certificate of participation is included with this workshop.
Instructor
Meet your instructor
Rajiv Goel
AI Transformation Leader and Microsoft Certified Trainer
Microsoft Certified Trainer with experience helping organisations translate Copilot and generative AI capabilities into practical workplace adoption.
FAQs
Common questions
Do I need coding skills?
No. The workshop focuses on agent design using accessible AI tools and low-code or no-code capabilities where available.
Which AI platform do we use?
The principles are tool-flexible. The hands-on build can be aligned to the approved platform available to participants.
What is the difference between an agent and a prompt?
A prompt is usually a single instruction; an agent packages recurring instructions, context and behaviour for repeated use.
Will I build an agent during the workshop?
Yes. Participants create and test at least one practical prototype.
What kinds of agents work well?
Repetitive tasks with clear inputs, expected outputs and review points are strong starting candidates.
Can I use company data?
Only in line with your organisation's policies and approved tools. Sanitised examples are recommended for open training environments.
How do I know if my agent is reliable?
You will create test scenarios and review outputs for accuracy, completeness and consistency.
Can I build more agents afterwards?
Yes. You leave with templates and an opportunity portfolio to continue building.
