How a 4-person team runs a 7-figure business with AI

The interface of work is changing

We used to click and type. Now we describe the outcome, and the AI does the clicking and typing for us. In this episode, brothers Cassy and Emil share how they built Postbeam, grew Hoppier, and ship faster by using AI to browse, ideate, and prototype.

This New Way is a weekly show and companion newsletter that turns real leader workflows into step-by-step playbooks you can run today.

AI tools mentioned in this episode:

• v0 for rapid UI prototyping

• Cursor for code comprehension and change impact

• Loom for quick context handoff

• Claude (Anthropic) for writing and structure

• 11Labs for quick text-to-speech

• Postbeam for multi-source curation and team posting

Tutorial 1: Turn a transcript into a LinkedIn post that sounds like you

Why it matters:
If you already speak on podcasts, host meetings, or join panels, you’re sitting on a goldmine of content. This workflow turns that raw material into polished posts that sound like you — without adding more meetings or writing time.

Step 1: Get your transcript

• If it’s a YouTube video, copy the transcript directly from YouTube.
  

• If it’s audio only, upload it any free transcription tool.
  

• If it’s from a meeting, export the transcript from your AI meeting notetaker (e.g. Fellow).

Step 2: Clean it up

Use AI to:

• Delete speaker names and filler words (“um,” “yeah,” “you know”).
  

• Keep only the parts with real insights or stories.

Timestamps are optional — only keep them if you want to quote moments later.

Step 3: Teach the AI your voice

• Find three of your past LinkedIn posts that sound most like your natural tone.
  

• Copy those posts into Claude (or ChatGPT) and ask AI to analyze your tone and voice.
  

• Then paste your cleaned transcript underneath.
  

• Ask it to write a single LinkedIn post that captures one clear idea, using your voice and tone.
  

• Keep it 120–180 words with a strong opening line that makes people want to click “see more.”

Step 4: Generate more ideas

• Once you like one post, ask the AI for eight more ideas from that same transcript.
  

• For each idea, have it list:
  

  • A one-line hook
    

  • A one-sentence summary of the post
    

• This gives you a month of content in minutes.

Step 5: Strengthen your hook

• Pick your favorite post and ask the AI to make the opening line stronger.
  

• Add a specific number or real result (for example, “We doubled engagement in 30 days”).
  

• Keep it true and believable.

Step 6: Post, measure, repeat

• Schedule one post per weekday for two weeks.
  

• Track impressions, comments, and saves to see what works.
  

• Keep your top two formats and repeat those styles next month.
  

Pro tip:

• If the post sounds robotic, paste in a few real emails or Slack updates you’ve written and tell the AI to match that tone.

👉 Watch the full walkthrough

Tutorial 2: Turn feature ideas into quick mockups your team can review

Why it matters:
Your sales, marketing, and support teammates hear customer feedback every day. Instead of explaining their ideas in long messages, they can now show them - by generating simple, clickable mockups.

Step 1: Create a shared space

• Upload your app’s current design (or component library) into v0 or a similar AI design tool.
  

• Share one team link in Slack so anyone can access it when they have an idea.
  

Step 2: Explain the goal, not the button

• When describing an idea, focus on what the user needs to do, not the specific UI.
  

• Example: “Let users schedule posts for later. They should pick a date and time, and see an error if they choose the past.”
  

Step 3: Let the AI mock it up

• The AI will instantly create a first draft of the feature.
  

• Then refine it in plain English:
  

  • “Make the time picker smaller.”
    

  • “Add helper text under the field.”
    

  • “Show a success message after saving.”
    

Step 4: Hand it off clearly

• Once it looks right, export the mockup.
  

• Record a 60-second Loom video explaining what it does and why it matters.
  

• Share both the mockup and video with your product or engineering team.
  

Step 5: Use AI for QA

• In Cursor (or a similar coding assistant), ask:
  

  • Which parts of the app does this change affect?
    

  • What tests should cover it?
    

• Use that as your QA checklist before release.
  

Pro tips:

• Give teammates a short “idea template” to fill out: problem, user, goal, and what success looks like.

• Treat these mockups as conversation starters, not final specs - product still prioritizes what gets built.

👉 Watch the full walkthrough

Bonus workflow to try next:

Use an AI-first browser like Comet to automate repetitive multi-click research. Describe the outcome, let it navigate, then export results to a sheet. Great for prospecting, marketplace searches, and price scans.

Why these playbooks matter: the fastest teams turn conversations and requests into shippable assets with near zero friction. That compounds.

👉 Subscribe to This New Way: YouTube Channel

 📩 Was this email forwarded to you? Subscribe here

📱Find us on Instagram and TikTok: @ This New Way

Until next time,

Aydin Mirzaee
CEO at Fellow.ai & Host of This New Way