Claude Cowork for Business

Course Details

A 2-day hands-on workshop, 6 hours per day (12 hours in total), delivered classroom or hybrid via Microsoft Teams. It emphasizes practice through workshops based on real organizational work, such as summarizing meeting reports, organizing files, building executive reports, screening job applicants and creating an interactive dashboard. Learners take home the Skills and templates they build in class (6-15 people per class, Intermediate level, no programming background required).

Day 1: Claude Fundamentals, Prompt Engineering and Business Applications

Section 1: Getting to Know Claude and the AI Agent Era

• The shift from chatbot to AI agent and its impact on organizational work
• Why Claude excels at documents, analysis and deep-reasoning work, compared with ChatGPT and Gemini
• The Claude model family (Opus, Sonnet, Haiku), choosing the right one, and limitations to be aware of

Section 2: Getting Started with Claude.ai Professionally

• The structure of Chat, Project, Artifact and Memory on both web and mobile
• Comparing Free, Pro, Max, Team and Enterprise plans and setting personal preferences
• Understanding tokens and the context window, plus a workshop: set up your profile and try Claude on your own real work

Section 3: Prompt Engineering for Management and Business Work

• The elements of a good prompt (context, role, task, constraints, output format) and role-setting techniques
• Chain-of-Thought, Few-Shot Prompting, System Prompts and iterative refinement
• Workshop: build a personal prompt library for 5 routine tasks such as meeting summaries, email drafts and proposal analysis

Section 4: Claude for Business Documents and Data Analysis

• Creating business documents (executive reports, meeting summaries, formal emails, proposals) and summarizing many long documents/PDFs at once
• Analyzing data from Excel and CSV, building Artifacts and interactive dashboards, and using Web Search and Deep Research
• Producing real files in Word, Excel, PowerPoint and PDF, plus a workshop: turn raw sales data into an executive report with charts and strategic recommendations

Section 5: Claude Projects - Build a Specialized AI Assistant

• What a Claude Project is, how it differs from a normal chat, and setting Project Instructions
• Uploading Project Knowledge (company handbooks, policies, reference documents) and example projects for marketing, HR, finance and sales
• Sharing a project with the team, plus a workshop: build a personal AI assistant for summarizing meeting reports

Section 6: Connectors - Link Claude to Your Data

• The Connectors concept and connecting Google Drive, Gmail and Google Calendar
• Access-permission and data-security precautions
• Workshop: connect Google Drive and have Claude summarize the key points from multiple documents at once

Day 2: Claude Cowork, Skills, Automation and AI Strategy

Section 7: Installing and Configuring Claude for Desktop

• The difference between Claude on the web and Claude for Desktop, and installing it on Windows and macOS
• Enabling the Virtual Machine Platform on Windows and fixing common issues
• The central Cowork settings: choosing the working folder and permission level, and verifying system readiness

Section 8: Getting to Know Cowork Mode and Agentic AI

• What Cowork Mode is and the agentic AI concept: plan, act, check and self-correct
• Multi-step tasks: issuing a long multi-step instruction and letting Claude carry on
• Requesting folder access safely, and reading/interpreting what Claude is doing, with stopping or redirecting midway

Section 9: Using Cowork to Manage Files and Produce Real Work

• Having Claude survey a folder, summarize its files, and categorize, rename and move many files by your rules
• Safety practices: backing up files, using a staging folder and confirming before deletion
• Workshop: organize a mixed simulated work folder and produce a report summarizing the whole folder

Section 10: Advanced MCP Connectors for External Tools

• What MCP (Model Context Protocol) is and an overview of usable connectors such as Slack, Notion, Google Workspace and task systems
• Installing, connecting and testing a connector safely, and producing submission-quality files (.docx, .xlsx, .pptx, .pdf)
• Workshop: build an automated workflow that pulls data from an external source, summarizes it and outputs a presentation file

Section 11: Skills - Work Standards Embedded in the AI

• What a Skill is, how it differs from a Prompt or Project, its structure, and writing descriptions so the AI invokes it at the right time
• The benefits of Skills (consistency, less training time, knowledge transfer) and creating, editing, testing and sharing them
• Workshop: build a "Corporate Brand Identity" Skill that forces every document to use the organization's colors, fonts and tone

Section 12: Applying Skills to Real Organizational Work

• How to choose work suited to a Skill (repetitive, high-volume, clear criteria) and examples in HR, sales, accounting and customer service
• Setting evaluation criteria and quality control, with precautions around bias and personal data
• Workshop: build a candidate-screening system from many resume files with an automatic scoring table

Section 13: Web Control and RPA with Claude in Chrome

• Installing and configuring the Claude extension on Google Chrome and having Claude open sites, read data and summarize
• Web scraping and automatic form filling, with legal and ethical precautions
• Workshop: an RPA Challenge - automatically enter data into a web form from a source data file

Section 14: Task Schedule, Live Artifact and Building a Landing Page

• What Task Schedule is, which tasks suit scheduling, and setting daily, weekly and one-time future tasks
• Live HTML Artifacts that fetch fresh data each time they open, and a 10-minute technique to build a landing page
• Workshop: build a personal work-tracking dashboard and schedule a task to summarize work every morning

Section 15: Capstone Workshop

• Brief: design and build one AI workflow for the learner's own real work, starting by identifying the most time-consuming routine task and breaking it into steps
• Design the Prompt, Project or Skill for the task, use Cowork to process real files and produce output files, then schedule a task or connect a connector to keep it running
• Present the work and estimate the time saved per month

Section 16: AI Strategy, ROI and Safe Practices for the Organization

• Assessing the value of AI, measuring time saved and basic ROI, and sequencing AI adoption in the organization
• Data-security practices, what data should not be entered into AI, and setting a basic organizational AI usage policy
• Reviewing AI output and keeping human responsibility (human in the loop), plus resources for further learning