Google Gemini Gems + Notebook LM: The Ultimate Free AI WorkflowDiscover how to combine Google Gemini Gems and Notebook LM to create an AI assistant that perfectly mimics your voice without ever hallucinating facts. Learn three powerful systems to automate your content writing, business proposals, and customer support in minutes.00:00 - Intro: The Ultimate AI Setup00:47 - What are Gemini Gems?01:13 - The PACK Formula for Instructions02:07 - Fact-Checking with Notebook LM02:37 - System 1: Content Writing Workflow04:32 - System 2: Proposal Writing Automation05:56 - System 3: Customer Support Systems07:11 - How to Get Started
Full transcript
Google Gemini Gems plus Notebook LM is insane. Today I'm going to show you a free AI setup that's going to blow your mind. It writes in your voice, it pulls real facts from your files, and it does it all in minutes. No more boring AI text.
No more made up stats. By the end you'll have your own AI helper ready to go. So here's the problem with most AI tools right now. You ask them to write in your voice.
They do a great job. But then you find numbers that don't exist. Stats that are made up. Facts from nowhere.
Or you go the other way. You upload your own files. The AI sticks to the facts. But it dumps it all in one big ugly wall of text.
And you spend an hour fixing the format. I hit this same wall a few weeks back. And I fixed it by linking two free Google tools together. Gemini Gems and Notebook LM.
When you put them together, something crazy happens. You get an AI that knows your stuff and writes how you want. Let me break this down for you. A gem is just a custom version of Gemini.
Think of it like a smart helper that never forgets your rules. You tell it who to be, what to do, the rules to follow, and how to format the answer. Then it remembers all that forever. To make one, head to Gemini and click Gems on the left.
Then click My Gems and hit New Gem. You'll see three boxes. Name. Description.
Instructions. The name is just for you. The description helps you remember what it does. The instructions box is where the magic happens.
Now here's the secret to making these instructions work. I use a simple system called PACT. Persona, Assignment, Context, and Template. Let me show you how this works for something real.
Say you run a coaching group like the AI Profit Boardroom. And you want to write welcome emails to new members. You'd build a gem like this. For Persona, you write, You are a friendly coach who helps business owners use AI to save time.
For Assignment, you write, For Context, you write, Always end with one clear next step. Match a tone that is warm but not too casual. For Template, you write, Save it. Now ask it to write a welcome email.
Boom. You get an email in the exact tone you wanted. Short. Clear.
With a next step. And here's the best part. Ask it to write 50 more. Every single one will follow those same rules.
You don't have to repeat yourself. Why does this matter? Because in a normal Gemini chat, your rules start to fade after a few messages. The AI starts to drift.
With a gem, your rules are locked in for life. But gems have one weak spot. They follow your rules well. But when you ask them to use real facts from your files, they sometimes guess.
They paraphrase loosely. They fill in the gaps with stuff that sounds right but isn't real. That's where Notebook LM comes in. Notebook LM is built to stick to your sources.
Every answer traces back to a real document you uploaded. It doesn't make stuff up. It can't. So when you put these two together, you get the best of both worlds.
Notebook LM keeps the facts true. The gem keeps the writing on point. Now let me show you the first system I built. Content writing.
Before this, I'd spend an hour reading articles. Pulling stats. Taking notes. Then I'd open a blank doc and start writing from zero.
The research never connected to the writing. I'd have to read my own notes again. Try to remember which stat came from where. Stitch it all together by hand.
Now here's how I do it. I open Notebook LM and create a new notebook. I upload a few sources. Maybe a report on AI tools for small business.
Maybe a few blog posts about AI automation trends. Then I switch the prompt box to deep research. I type something like, find more sources on how small business owners use AI automation to grow in 2026. Notebook LM digs through dozens of sources.
When it's done, I hit import. Now I have a full research base inside the notebook. I go to the studio panel on the right. Click reports.
Pick create my own. Notebook LM gives me a clean research brief. Every claim links back to a real source. I can check anything I want.
Then I export this to Google Docs. It saves to my drive. Now here's where the gem comes in. I make a gem called content writer.
I write my full packed formula in the instructions. My voice. My tone. My format rules.
I don't upload any specific research to the gem. The gem just holds my style. That way I can use it for any topic forever. I open Google Docs.
Click the Gemini icon in the top right. The sidebar opens. I hit the plus icon and turn on drive search so it can read my files. I pick my content writer gem.
Then I type. There's a research brief in my drive called AI Automation Trends 2026. Read it and write a 1200 word blog post for the AI profit boardroom. Show how AI automation helps small business owners save time and get more leads.
Follow my voice and structure. The gem pulls the report from my drive. Reads the real research. Writes a full blog post in my exact voice.
The whole thing takes about 15 minutes. Start to finish. Now I want to stop here for a quick sec. If you're creating all this content with Notebook LM but nobody's finding it online, you've got a traffic problem.
That's where SEO comes in. If you want more leads, more traffic, and more customers, my team can help you rank higher on Google. Check the link in the comments and description to book a free SEO strategy session. Let us handle your backlinks while you focus on growing your business.
Okay, back to the systems. The second one saves me a ton of time on a different task. If you run any kind of service business, you know this pain. Every new lead means a new proposal.
And every proposal is 80% the same as the last one. You dig through old projects. Copy the parts that worked. Change the scope.
By the time you're done, an hour is gone. Here's how I cut that to 15 minutes. I open Notebook LM and make a notebook called Business Knowledge Base. I upload everything that defines my business.
Case studies with real results. Service descriptions. Then I go to the studio panel and make a report. Notebook LM pulls it all into one clean reference doc.
My services are organized. My case studies are summed up with the client type and results. I save it to Google Drive. Now I make a gem called Proposal Writer.
I add the instructions for how I want proposals to look. When a new lead comes in, I open Google Docs. Click the Gemini icon. Turn on Drive Search.
Pick my proposal writer gem. I type something like, there's a master reference doc in my drive called Business Knowledge Base. Use it to write a proposal for a new AI profit boardroom partner who wants to add AI automation to their team. Timeline eight weeks.
They care about saving time and getting better results from their marketing. The gem pulls my reference doc. Reads my services. Finds two case studies that match.
And drafts a full proposal with the project scope, timeline, and case study links. Without Notebook LM in the mix, the gem would guess at details and make up case studies. With the pre-organized reference, it pulls real projects. And I can share this gem with my team.
So anyone on my team can write proposals using the same rules and the same facts. Without me being involved. Now the third system handles something even more time consuming. Customer support.
If you run any kind of business, the same questions come in every day. Where's my order? Can I get a refund? How do I cancel?
And every answer needs to match your real policies. It needs to sound like your brand. It needs to be the same no matter who on your team is replying. If one person says 30 days for returns and another says 14, that's a real problem.
Here's how I fixed it. I open Notebook LM and make a notebook called support knowledge base. I upload all my support docs. Queue page.
Return policy. Shipping policy. Subscription terms. A brand voice guide.
Then I generate a briefing doc from the studio panel. This squeezes my docs into one clean reference. The main rules. The edge cases.
The brand voice notes. I also generate a response guide by question type. I save both to Google Drive. Then I make a gem called support helper.
I write the instructions. I open Google Docs. Click the Gemini icon. Turn on Drive search.
Pick the gem. Now I paste a customer message. Like I joined the AI profit boardroom three weeks ago and I want to know how to cancel my membership. The gem pulls my policy docs from Drive.
Finds the right cancel rule. Matches my brand voice. And drafts two replies. A full email.
And a short chat reply. Same rules every time. You can use this same system for promo emails. Social posts.
Any reply your team sends out. One gem becomes a full company support system. New team members can be ready in a day instead of weeks. Because the gem already knows every policy.
So here's what to do next. Pick one of these three systems. The content writer. The proposal writer.
The support helper. Build it this weekend. It takes about 20 minutes. And once you have one working you'll see how easy the others are to set up.
The whole point is simple. You stop wasting time on stuff a machine can do. You spend that time on stuff that actually grows your business. That's the real win.
And one more thing. If you're creating all this content with NotebookLem but nobody's finding it online you've got a traffic problem. That's where SEO comes in. If you want more leads, more traffic and more customers my team can help you rank higher on Google.
Check the link in the comments and description to book a free SEO strategy session. Let us handle your backlinks while you focus on growing your business. Thank you for watching. Drop a comment below and let me know which system you're going to build first.
I'll see you in the next one.
More episodes