感謝您提交詢問！我們的一位團隊成員將在短時間內與您聯繫。
感謝您提交預訂！我們的一位團隊成員將在短時間內與您聯繫。
課程時長 14 時間：
課程簡介
- Day 1: Understanding Claude and using it in daily work
- Arrival and target picture
- Current status of each participant and available licenses
- Each participant brings one use case to work on for two days
- What Claude can do today and where its limits are
- Exercise: Write the own use case on a card with goal, input material and desired result.
- How language models work
- Tokens and context window, meaning how much text a model considers at once
- Why the same question can produce two different answers
- Why models invent details and how to contain this
- Model matrix of Haiku, Sonnet and Opus by quality, speed and usage limit
- When Claude, Copilot or ChatGPT is the better choice
- Exercise: Solve the same task with two models and compare the results by quality and speed.
- Good instructions and context
- Building blocks of a good prompt: role, task, context, examples, format and limits
- Letting Claude ask clarifying questions before it starts
- Summarizing and comparing long documents and requesting supporting passages
- Building review steps directly into the instruction
- Exercise: Improve a weak prompt from the own daily work in three rounds and compare the versions side by side.
- Projects as shared memory
- Splitting project instructions and knowledge files sensibly
- Storing house style, product knowledge and templates centrally
- Separating team and private areas and keeping sensitive content apart
- Checking, correcting and, if needed, switching off memory
- Exercise: Create a project for the own use case with an instruction and three knowledge files.
- Artifacts, files and analyses
- Reports, tables, charts and presentation drafts as artifacts
- Querying and comparing Excel, PDF and Word files
- Letting the analysis tool calculate figures and spot-checking them
- Exporting, publishing and sharing artifacts in the team
- Exercise: Create an analysis with a chart and a short comment from an anonymized Excel list.
- Outcome: Each participant has a set-up project for their use case, two improved templates and a first artifact built from real material.
- Arrival and target picture
- Day 2: Extending Claude with skills, connectors, Cowork and routines
- The ecosystem at a glance
- Where things run: chat, desktop app, Claude in Chrome, Cowork and Claude Code
- What a skill, a connector, MCP and a routine are
- Agents explained simply: a workflow with AI and tools
- Live demo from idea to finished skill
- Exercise: Assign the own use case to the matching layer and justify the choice.
- The first custom skill
- Structure with description, when the skill applies, workflow, examples and templates
- Writing and improving skills together with Claude
- Testing with three real cases
- Sharing and maintaining skills in the team
- Exercise: Each participant builds a skill for their use case, for example replies in the house style or a review report.
- Connectors and connected services
- Connecting Microsoft 365 or Google Workspace, file storage and calendar
- Keeping permissions deliberately small
- Recognizing hidden instructions in emails and documents (prompt injection)
- Exercise: Connect a connector and solve a task across mailbox and calendar, for example preparing the next appointment.
- Claude Cowork and routines
- Handing over tasks instead of chatting, with access to a released folder
- Reading the plan, following progress and reviewing interim results
- Scheduled tasks, for example a morning overview or a weekly report
- Approval points before anything is sent or deleted
- Exercise: Set up a routine for a daily overview and evaluate the first run together.
- Team rules and the next 30 days
- What may be uploaded and what may not
- Maintaining shared skills and projects and assigning responsibilities
- Model matrix and reference sheet as a handout
- Exercise: Write a personal 30-day plan with three concrete steps.
- Outcome: Each participant leaves the course with a working skill, a connected connector, a set-up routine and a plan for the next 30 days.
- The ecosystem at a glance
最低要求
- Laptop with a current browser, Claude account (Pro, Team or Enterprise), Claude desktop app recommended, and a real use case from daily work.
- No programming skills required.
Audience
- Employees from administration, finance, purchasing, HR, project and process management who use or are introducing Claude in their company and want to go beyond the chat.
客戶評論 (3)
全面了解背景；如何进行提示。
Olaf Kannt - Helios Klinikum Krefeld
課程 - Generative AI and Prompt Engineering in Healthcare
機器翻譯
本次培训中可借鉴的技巧与推荐提示
Lee Mei Lin - ST Engineering IHQ Pte Ltd
課程 - InVideo AI: Creating Engaging Short-Form Videos
機器翻譯
使用正确且有效的提示词
Marses Pacaldo
課程 - Generative AI and Prompt Engineering for Corporate Professionals
機器翻譯
