Build Business Apps with Claude Code

Day 1: Start Building Apps with Claude Code and Lay the Project Foundation

Section 1: Kickoff and Mindset Shift

• How software development is changing in the AI coding agent era, and what Vibe Coding is versus traditional programming
• Shifting from coder to Creative Director who directs the outcome, with a mindset reset: say what you want, not how to write it
• Demo: build a working web app in 5 minutes, and where AI coding excels versus work still needing human judgment

Section 2: Installing and Configuring Claude Code

• Getting to know Claude Code and its three usage channels (Terminal, Claude Desktop, VS Code)
• Hands-on: install Claude Code on your machine, choose a subscription plan (Free, Pro, Max, Team/Enterprise) and verify the install
• Understanding the working directory, granting the AI file access, common commands and reading the AI's responses

Section 3: Building Your First Application

• A simple project structure and the difference between Plan Mode and Auto Mode and when to use each
• Reviewing and approving the plan before the AI acts, and the feedback loop: instruct, review, fix, repeat
• Viewing the result as a working app in the browser, plus a workshop: build a personal dashboard such as a to-do list, expense logger or calorie tracker

Section 4: Prompt Engineering for Building Apps

• The structure of an effective prompt: Context + Goal + Constraints + Examples, plus context stuffing and specification by example
• Incremental prompting instead of one big instruction, and using mockups, screenshots and attachments with your prompt
• Common pitfalls and how to fix them, plus a workshop: improve the Section 3 app with quality prompts

Section 5: CLAUDE.md and Project Memory

• What CLAUDE.md is, why it matters for output quality, and turning a note file into a work contract the AI must follow
• Standardizing the project structure (src/, assets/, docs/ folders and naming conventions) and defining coding style, communication language and prohibitions
• The memory tier concept (working memory vs knowledge base), global vs project CLAUDE.md, plus a workshop: write a CLAUDE.md for your project and test before and after

Day 2: Move to Real Use, Connect Data and Deploy Online

Section 6: Skills and Slash Commands

• Overview: Skills vs Slash Commands vs CLAUDE.md and when to use each, plus the structure of a slash command to build personal commands such as /summarize, /translate
• The structure of a Skill (SKILL.md, a good description, supporting files) and writing descriptions so the skill is invoked at the right time
• Managing and sharing skills with teammates, plus a workshop: build a weekly-report skill that reads data from a file and summarizes it into an automatic report

Section 7: MCP - Connecting Claude Code to Real Data

• What MCP (Model Context Protocol) is, what it solves, and a brief architecture (MCP Server, MCP Client, Tool)
• Installing and connecting common MCP servers (Google Drive, Notion, GitHub, Database) and verifying the connection
• Workshop: pull data from a Google Sheet into your app, with security around permissions, scope, token handling and precautions for personal and confidential data

Section 8: Persisting Data with Supabase

• Why a business user should know a ready-made backend, and the difference between browser data and real database data
• Creating a Supabase project and designing tables by instructing the AI, connecting the app to Supabase and testing writing/reading data
• Basic access-control concepts (Row Level Security), plus a workshop: make the app store data in Supabase for use across devices and sessions

Section 9: A Professional Development Process with AI

• The 5-step process - Interview, Plan, Build, Test, Ship - and having the AI interview your requirements before starting
• Using sub-agents by role (Researcher, Reviewer, Tester) and the Hooks concept for automatic quality control
• Designing a modular structure, debugging approaches for non-programmers, and keeping the AI from changing things you did not ask for

Section 10: Capstone Workshop - Build a Tool to Solve a Real Problem

• Brief: pick a problem from your own daily work and build an internal tool or web app such as an expense tracker, lead qualifier, meeting summarizer, meeting-room booking or task tracker
• Steps: scope it and write a CLAUDE.md, plan with Plan Mode, build and refine with the feedback loop, connect a Supabase database and test and fix errors
• Present the work and exchange approaches among learners

Section 11: Deploy, Share and Real Organizational Use

• Deploying to Vercel or Netlify in minutes and sharing the link with colleagues or customers
• Basic version control for non-programmers (Git and GitHub) and rolling back to a previous version when the result is not right
• Estimating the cost of AI and related services, security and organizational-policy considerations, when to hand off to a professional dev team, and paths for further learning