Claude Code for Business Automation

Course Description

This practical course introduces business professionals to building automations with Claude Code. Designed for non-developers, it shows how to turn everyday business requirements into working solutions using AI-assisted programming, often called “vibe coding”. Participants learn how to identify suitable processes for automation, write clear requirements and guide Claude Code through planning, building and testing a solution. The course covers working with files, managing permissions and costs, connecting business tools and deploying automations with appropriate human oversight. An expense receipt case study develops throughout the course, progressing from simple data extraction to a complete workflow using multiple AI agents and human review. Participants also develop their own automation project, leaving with a working prototype and practical experience they can apply in their organisation.
5 Days
Contact us for pricing
 

Who Should Attend?

Team leads and managers exploring automation within their function
Business analysts and operations professionals working with repetitive processes
Professionals responsible for scoping or championing automation initiatives
Non-developers who want to build practical business solutions with AI

Prerequisites

No programming background is required. Programming concepts and the tools needed to build automations are introduced throughout the course.

Introduction to Claude Code

Understanding AI agents and how they differ from chatbots
What Claude Code can and cannot do
Identifying suitable business processes for automation
Understanding where human judgement is required
Installing Claude Code and understanding pricing

The Working Environment

Introduction to Visual Studio Code
Using the integrated terminal
Understanding files, folders and paths
What is code, a script and an application?
Introduction to dependencies and libraries
Reading error messages and basic debugging

Working with Files and Prompts

Giving Claude Code access to project files
Reading and editing files directly
Extracting information from images and PDFs
Understanding tool use
Prompting techniques for business tasks
Extracting structured data from scanned receipts

Planning and Specifying an Automation

Using CLAUDE.md to provide persistent project instructions
Turning a business requirement into a written specification
Breaking a solution into manageable features
Writing testable requirements
Using Plan Mode to review work before execution
Defining business policies and rules

Managing Context, Sessions and Costs

Understanding the context window
Tokens and their effect on cost
Choosing a model for the task
Using /compact and /clear
Managing longer working sessions
Keeping project context relevant and usable

Permissions and Human Oversight

Understanding permission modes
Controlling file access and permitted actions
Defining which decisions can run automatically
Setting approval and escalation thresholds
Adjusting effort levels
Interrupting and redirecting work
Managing memory and project versus global settings
Using slash commands for repeatable tasks

Reusable Workflows and Automation

Packaging repeatable workflows as Skills
Using Hooks to trigger actions
Incorporating web browsing into a workflow
Building a reusable receipt extraction process
Connecting workflow steps into an automation

APIs and Business Tool Integrations

What is an API?
Understanding requests and responses
Introduction to GET, POST and JSON
API keys and authentication basics
Connecting tools through Model Context Protocol (MCP) servers
Worked examples using DeepWiki and Perplexity MCP
Connecting workflows to business tools such as Google Sheets and Slack

Working with Multiple Agents

Introduction to subagents
Dividing a process into clearly defined tasks
Coordinating agents within a workflow
Separating extraction, categorisation, authorisation and auditing
Applying business rules consistently
Routing outcomes for approval, rejection or human review

Version Control and Deployment

Introduction to Git version control
Versioning policy files and automation logic
Running locally versus deploying an internal application
Introduction to dashboards, forms and user access
Cloud deployment and preview environments
Introduction to Remote Control
Deploying to a live server using SSH
Understanding SSH keys and credential handling
Verifying a deployment and documenting the solution

Running and Maintaining Automations

Scheduling unattended tasks
Logging and monitoring
Handling errors and sending notifications
Planning backups
Assigning ownership and support responsibilities
Maintaining and updating the solution
Monitoring ongoing costs

Practical Project: Expense Receipt Automation

Extracting receipt data into a structured format
Categorising expenses against a written policy
Checking claimant roles and spending limits
Auditing claims against defined risk and escalation rules
Recording approval and rejection decisions
Notifying a human reviewer when a claim requires attention
Incorporating human decisions into the workflow
Deploying and scheduling the completed pipeline

Participant Automation Project

Identifying a business process to automate
Developing an individual project alongside the taught sessions
Applying course techniques during six days of independent project work
Producing a working automation prototype
Demonstrating the solution and explaining its workflow and human review steps
Artificial IntelligenceClaudeAnthropicAI