Course Outline
Introduction
-
Defining business automation using ChatGPT
-
Models, agents, tools, and workflow structures
-
Fixed workflows versus agentic processes
-
Selecting the appropriate level of autonomy
-
The importance of human oversight and decision-making
-
Moving beyond single conversations: reusable workflows and recurring executions
Mastering Context and Memory
-
Chat context versus persistent memory
-
Storing process state explicitly
-
Maintaining decisions and key information across multiple runs
-
Context limitations and potential loss of earlier data
-
Continuing work across multiple conversation sessions
-
Distinguishing between permanent rules and current process status
Organizing Work in ChatGPT
-
Risks associated with combining excessive tasks and changing requirements in one thread
-
Identifying missing decisions and inconsistent outputs
-
Breaking down work into smaller, well-defined tasks
-
Transferring goals, sources, decisions, and results between tasks
-
Determining when to use separate conversations or workflows
Utilizing Plugins, Projects, and Spaces
-
The purpose of plugins, projects, and Spaces
-
Organizing reusable instructions and tools
-
Linking workflows to business information sources
-
Managing related conversations and shared materials
-
Sharing documentation and results effectively
-
Preparing the working environment for business processes
Constructing Workflows with Memory
-
Setting the workflow objective
-
Defining inputs, steps, outputs, and acceptance criteria
-
Recording process state between executions
-
Applying previous decisions in subsequent runs
-
Managing errors and incomplete executions
-
Resuming workflows after interruptions
-
Developing a prototype workflow for a chosen business task
Scheduling and Consistent Execution
-
Manual versus recurring workflow execution
-
Setting execution frequency and time zones
-
Establishing notification and stopping rules
-
Managing scenarios where new data is unavailable
-
Re-running workflows with saved state
-
Comparing outcomes from repeated executions
Assessing Workflow Quality and Consistency
-
Establishing quality criteria
-
Verifying required fields, numbers, sources, and output formats
-
Detecting duplicate or inconsistent results
-
Differentiating acceptable wording variations from process errors
-
Understanding sources of variability in AI-generated content
-
Testing workflows using representative business cases
Practical Workshop
-
Selecting a recurring business task
-
Designing the workflow
-
Defining process memory and state
-
Executing and testing the workflow
-
Analyzing multiple executions
-
Identifying errors and areas for improvement
-
Preparing the automation for real-world use
Troubleshooting
-
Missing or incomplete context
-
Incorrect or outdated process state
-
Inconsistencies between executions
-
Missing data or unavailable sources
-
Duplicate processing issues
-
Failed or partially completed workflow runs
-
Limitations in tool or account features
Summary and Future Steps
-
Reviewing the completed workflow prototype
-
Identifying tasks suitable for automation
-
Defining quality and acceptance criteria
-
Planning integration into daily work
-
Exploring opportunities for an advanced second-day module covering code-supported processing and agent development
Requirements
-
Recommended Knowledge Before the Course
- Familiarity with ChatGPT basics
- No programming experience is required as a prerequisite.
- Participants should possess a computer, an internet connection, and a ChatGPT account that includes access to the features utilized during the workshops.
- Ensure plugins, Spaces, and scheduling/automation features are accessible before the session, as availability may vary based on account or organizational settings.
Target Audience
This course is designed for:
- Specialists and managers integrating ChatGPT into their daily operations
- Process owners and business analysts tasked with optimizing team workflows
- Professionals responsible for creating reports, summaries, documents, data compilations, and recurring business updates
- Team leaders deploying AI and coordinating work through shared information sources
A concise summary for catalogues:
Prerequisites: Basic understanding of ChatGPT. No coding background necessary.
Audience: Managers, specialists, process owners, business analysts, documentation professionals, and AI implementation leaders.
Testimonials (1)
Able to pivot upon audience suggestions - ie able to create a real AI agent scenario on the spot.
