Finish the meeting with less work left to carry, Complete Resource Pack

Ideas to keep

1. 01A useful answer to a client often needs preparation first: the notes from the last meeting, the missing facts from your team, and a plan if the date is at risk.
2. 02AI helps with that preparation when you give it a clear job and the notes it needs.
3. 03Software that works through several steps toward a result on its own, and saves one document for you to inspect, is called an AI agent.
4. 04A quick chat reads what you attach and answers once. An agent keeps going until the finished document is saved for you to check.
5. 05A polished draft can still miss the point. Your knowledge of the work closes that gap, so you examine, question and improve the plan before you speak.

Try it (optional)

1. Think of one client or team question in your week that needs preparation before you can answer it.
2. List what that preparation involves: notes to find, facts to chase, a plan to draft.
3. Notice which of those steps you would want an agent to do, and which decision stays yours.

Files and links

You do not need coding experience for any strategy in this course. Practice is always optional.

• ChatGPThttps://chatgpt.com
• Geminihttps://gemini.google.com
• Download the Section 1 starter materials (zip)/downloads/leaders/section-01/section-01-starter-materials.zip

Ideas to keep

1. 01Priya's team can start account setup on Tuesday only after the client sends a role list showing who needs access and what each person may do.
2. 02A standard summary can repeat everything Jordan said and still leave Alex chasing updates before Monday.
3. 03The transcript is the written record of what people said. Alex's purpose for Monday tells the agent what to look for in it.
4. 04A blocker is something preventing the work from continuing. Asking what remains open gets a more useful reply than asking for a status.
5. 05Every message stays a draft and the reminder stays a proposal until Alex approves it, because the person whose name is on the message checks it.

Try it (optional)

1. Open practice/meeting-transcript.md and find Jordan's line promising the role list by 3 p.m.
2. Open practice/meeting-follow-through.html. On the Meeting notes tab, find the client role list under Waiting for confirmation, then Alex's decisions.
3. Switch to Follow-up drafts and read the draft to Jordan. Find the sentence that asks what remains open if confirmation is delayed.
4. Open the Reminders tab and check that the 10 to 10:15 a.m. Monday reminder says Proposed only.

Files and links

Northstar is a prepared fictional example written for this course. The follow-through page shows the target an agent works toward; it is not a recording of a live agent run.

• Northstar follow-through, the prepared example/resources/library/leaders/northstar-follow-through-prepared-example
• Northstar case files: meeting transcript and leader context/resources/library/leaders/northstar-meeting-transcript-and-leader-context
• Open the interactive follow-through page/downloads/leaders/section-01/meeting-follow-through.html

Ideas to keep

1. 01You do not need a polished written request to begin. You can turn on the microphone and describe what is bothering you.
2. 02The instructions you give AI to start or guide a job are called a prompt. A prompt can begin as an unfinished spoken explanation.
3. 03Asking AI to interview you first gives it a problem to investigate before it prepares anything.
4. 04Three questions were enough here: what Alex discovers too late, what Alex already has, and what each person should send back.
5. 05When you already know the exact output, prompt directly. When the approach is still unclear, a short spoken interview helps you find it.

Try it (optional)

1. In ChatGPT, turn on the microphone and describe a meeting problem in your own words, or read Alex's account aloud.
2. Add the instruction to interview you before starting the work.
3. Answer each question AI asks. Compare your answers with Alex's in prompts/section-01-prompts.md.
4. Judge the plan AI proposes: does it address your real worry and use the material you have?
5. When the plan fits, confirm it and ask for the page back for your review.

Files and links

The interview exchange is prepared for the course. Your own tool will ask different questions; judge them by whether they bring out what you need.

• Northstar case files: meeting transcript and leader context/resources/library/leaders/northstar-meeting-transcript-and-leader-context

Ideas to keep

1. 01Polite and professional is not the test. What matters is what the reply lets you learn before your next decision.
2. 02A generic progress request can bring back a list of completed tasks and still leave Alex needing another message.
3. 03A purposeful request asks what is preventing progress and whether a decision is needed, so the answer arrives while Tuesday's options are still open.
4. 04The agent reads, sorts and drafts. Alex supplies the purpose and judges whether the drafts ask for the right information.
5. 05Ask of any AI draft: does it merely sound finished, or does it set up the next decision I have to make?

Try it (optional)

1. Open practice/follow-up-exercise.md and read the meeting excerpt.
2. Compare the generic request with Alex's purposeful request.
3. Try revising the generic request yourself, or with AI, then open the sample answer and compare.

Files and links

• Improve one follow-up, the optional exercise/resources/library/leaders/improve-one-follow-up-exercise

Ideas to keep

1. 01Generative AI is the ability to create new wording, such as the sentences in Jordan's draft.
2. 02The model is the part that reads, reasons and writes. In ChatGPT, that is one of OpenAI's GPT models.
3. 03A tool is an ability to take an action, such as reading a file or saving one. Even a chat that reads an attached file is using a tool.
4. 04An AI agent combines the model and its tools until the job is finished, and can take only the actions its tools allow and you permit.
5. 05Agentic AI describes how it works: choosing and adjusting its own next steps toward your goal, such as asking when the purpose is missing.
6. 06A rule is one short instruction. A skill is a saved set of rules, with examples or supporting files, that the agent can reuse.

Try it (optional)

1. Open skills/meetings-and-follow-through/SKILL.md from the section zip and read its table of steps.
2. Find where it hands work to smaller skills such as meeting-summary and meeting-follow-up.
3. Open references/working-contract.md and find the line saying a prepared draft is not an executed action.

Files and links

The skill files are authored course materials. They have not been run in a live agent yet, so check how your own tool handles them before relying on them. In ChatGPT, agent work happens in a mode called Work, which Section 2 opens.

• Meetings and follow-through workflow/resources/library/leaders/meetings-and-follow-through-workflow

Ideas to keep

1. 01Assigning every meeting task at once mixes different people, files and judgment calls, and correcting the result starts to feel like a second job.
2. 02One review's follow-through is a clear first job: the transcript exists, the purpose is known, and the drafts are easy to judge.
3. 03A brief is a short description of the job, its inputs and its return point. AI can draft it from your conversation.
4. 04Check the brief: the inputs exist, the output is concrete, and it stops and returns for review without sending anything.
5. 05A workflow is a repeatable sequence of steps for a bounded job. Keep it if the drafts needed only small corrections; change the brief if you still had to chase people.
6. 06Four sizes: a rule is one instruction, a brief is one job, a skill saves a method, a workflow is the whole repeatable sequence.

Try it (optional)

1. Pick one recurring meeting in your own week.
2. Ask AI to turn the plan you agreed into a short job description.
3. Make the 3 checks: inputs exist, output is concrete, it returns for review.
4. After the meeting, decide whether to keep the workflow or change the brief.

Ideas to keep

1. 01Follow the lessons straight through to understand each strategy, then try it when the right meeting or decision arrives.
2. 02Return by problem: the interview lesson when you are unsure what to ask, the first-workflow lesson when you are tempted to assign too much.
3. 03The section files are fictional and ready to use, so you can practise without gathering confidential workplace documents.
4. 04When you apply a method to your own work, use the documents and AI products your employer permits.
5. 05Rough, unfinished questions are welcome in Q&A. Share the prompt you used and a non-confidential part of what came back.

Try it (optional)

1. Download and unzip the Section 1 starter materials.
2. Open README.md, then the practice and prompts folders.
3. Use captions or playback speed in the player whenever that helps you review a lesson quickly.
4. Post a question in the course Q&A when an explanation feels unclear.

Files and links

The zip holds the meeting workflow and the 8 skills it uses. The complete free resource pack, with all 3 starter workflows, comes later in the course.

• The course on Udemy (Q&A, captions and lectures)https://www.udemy.com/course/gen-ai-leaders/?referralCode=B6E0DB8B8D15355092C0
• Download the Section 1 starter materials (zip)/downloads/leaders/section-01/section-01-starter-materials.zip

Ideas to keep

1. 01You do not have to write the preference yourself. Ask the agent to capture what made the follow-up useful and why.
2. 02A preference is one short rule written with its reason. A skill can hold several rules like it, along with examples.
3. 03Northstar's Monday, October 5, 11 a.m. review belongs in Northstar's own notes, not in a general instruction.
4. 04Test a preference by applying it to a different task, such as asking a colleague for a cost figure before a budget decision.
5. 05Right now the preference lives only in Friday's conversation. Keeping it ready for next week is Section 2.

Try it (optional)

1. Open the Useful preference tab in practice/meeting-follow-through.html.
2. Read the three sentences of the proposed preference and judge whether they keep the reason the follow-up worked.
3. Find the meeting fact kept separately in the project record.
4. Apply the preference to one request in your own week and check that the action and reason are clear.

Files and links

• Northstar follow-through: the useful preference/resources/library/leaders/northstar-follow-through-prepared-example#preference