Everyday AI Workflows

Overview

1. Meeting recap

Tools: an AI notetaker plus a general assistant.

1. Connect the notetaker to your calendar so it joins meetings automatically.
2. Let it record, transcribe, and produce a first summary.
3. Paste the transcript into an assistant for a polished recap.

From this transcript, give me a 3-line summary, a table of action items (owner | task | due date), and a short follow-up email to attendees.

2. Rehearsal by voice

Tools: any assistant with a voice mode.

1. Describe the role, the organization, and the kind of conversation.
2. Have it question you out loud, one question at a time.
3. Ask for feedback after each answer: clarity, confidence, what to cut.

Act as a hiring manager for a product role. Interview me one question at a time.
After each answer, give brief, honest feedback before the next question.

3. Deep research

Tools: a citation-based answer tool or an assistant's deep-research mode.

1. State the question precisely: scope, and what a good answer contains.
2. Let it gather sources, then skim the citations for credibility.
3. Fill gaps with follow-ups, then ask for a structured summary.

Research <topic> for a business decision. Cover the current state, the main options with trade-offs, and a recommendation, with citations I can check.

4. Dashboard decoding

Tools: any assistant that accepts images or file exports.

1. Upload a screenshot or the underlying export.
2. Ask what the key numbers mean and what has changed.
3. Ask for the one or two things that deserve action.

Here's a screenshot of our analytics dashboard. Explain what it says in plain
English, highlight anything unusual, and tell me the two things to act on.

5. Spreadsheet analysis

Tools: a general assistant, or an in-spreadsheet AI add-on for row-by-row work.

1. Upload the file (or describe the columns) and state your question.
2. Ask for trends, outliers, and a plain-language reading.
3. Request a chart or summary table for your report.

Here's our sales data. Find the top 3 trends, flag anything surprising, and suggest what to investigate next. No jargon.

6. Recurring content

Tools: a general assistant with a reusable project that holds your voice and examples.

1. Give the audience, goal, and voice (an example helps a lot).
2. Draft, then refine with specific feedback: "shorter," "warmer," "stronger hook."
3. Save the winning setup as a project or skill so next time starts from your standard.

Using the voice guidelines in this project, write a short social post about <topic>: a hook in the first line, 3 short insights, and one takeaway. No hashtag clutter.

7. Ask your own documents

Tools: an assistant with file upload or connectors, or a tool built to answer only from your sources.

1. Upload files or connect the source.
2. Ask questions answered strictly from those documents.
3. Ask for summaries, comparisons, or an audio overview.

Based only on the documents I've uploaded: what are the key obligations in this contract, and where exactly does each one appear?

8. Act on a web page

Tools: a browser-sidebar assistant.

1. Open the page: an article, long doc, or dashboard.
2. Ask the sidebar assistant to read and summarize it.
3. Ask it to extract data, draft something, or take the next step.

Summarize this page in 5 bullets, put any dates and figures in a table, and draft a 3-line note I can send my team.

Starter stacks by role

Start small: two or three tools you actually use.

Previous: Working with an AI AssistantNext: AI Coding Agent CLI

Educational and general, not legal advice. Follow your organization’s policies and applicable law.