Hey
I used to treat AI tools like apps on my phone.
Download one. Use it hard for a week. Forget it existed by the next Monday. My home screen turned into a graveyard of tools I was sure would change my life.
The problem was not the tools. It was that nobody tells you which ones are hype and which ones actually pull their weight.
There is so much noise right now that picking randomly feels like the only option.
So I did the boring thing. I tested hundreds of them. I trained professionals on them. And then I cut the list down to seven.
Seven tools. They power about 90 percent of my work. And together they have saved me hundreds of hours I used to lose to busywork.
Here is the exact stack, and what each one is actually for.
1. The chatbots (and which one for which job)
Here is the assumption I have to kill in almost every training I run. People think [ChatGPT](https://www.chatgpt.com), [Gemini](https://www.gemini.google.com), and [Claude](https://claude.com) are basically the same, so it does not matter which one you grab. That is wrong, and using them randomly is a bad strategy.
Each one has a job.
Claude is your writer. It produces better first drafts, the writing sounds more human, and you need fewer revisions. Feed it samples of your own work and it matches your voice, so the output actually sounds like you.
I use it for presentations, emails, and proposals, anything where tone matters as much as content. If you vibe code, it also writes working code on the first try more often than anything else I have used.
Gemini is your Google brain. If you live inside Google Workspace, this is the one. It has a huge context window, so you can drop in a long video, a giant slide deck, or a massive client request and it handles all of it. It also digs through long Gmail chains and buried Drive files to find that one thing you half remember.
Pick the right chatbot for the right job and you get the right result the first time. That alone saves hours of redoing work.
2. [Perplexity](https://www.perplexity.ai/) (your accuracy layer)
Perplexity is the specialist I pair with my everyday chatbots.
It is built to find accurate, source based information, the kind that general models sometimes hallucinate.
So when my work needs verified facts, like a client research report or background for an interview, Perplexity becomes a second layer of checking on top of whatever ChatGPT or Claude gave me.
One tip. If you want to limit your research to a single source, you can prompt it to narrow down to just that one source. It keeps the answer clean and grounded.
3. [Whisperflow](https://wisprflow.ai/) (talk instead of type)
I will admit it. I ignored voice to text for years. Every version I tried felt like more trouble than it was worth, and I always ended up back at my keyboard.
Then I tried Whisperflow out of pure desperation. I wanted to fire off a reply to my team while chasing my very active daughter around the playground.
The math is what got me. We speak around 200 words a minute and type around 50. That means I produce writing four times faster just by talking.
Notes, replies, even AI prompts that would have taken forever to thumb out.
The part I love is that it does not just transcribe, it cleans as it listens. Change your mind mid sentence, fumble a word, or say "um" ten times while thinking out loud, and it figures out what you meant. What comes out the other side is clean, coherent text. Now I cannot get through a day without it.
Gamma was built by a founder who was not a design person stuck making corporate presentations. That is me. That is most of us.
There is a design tax most people pay without realizing it. You build a great pitch, the design is weak, and the idea does not land the way it should, because we judge the quality of thinking by how it looks on the slide. That is just human nature.
Great design takes time, skill, and tools most people were never trained on. Gamma removes all three.
I hand it my website and branding, and it designs a full, consistent, genuinely good looking deck, faster and better than I could do by hand. If you have ever lost an evening to slide formatting, this one buys it back.
5. [NotebookLM](https://notebooklm.google/) (research that never lies to you)
Here is something nobody talks about enough. We live in the most information rich moment in history, and most of us are drowning in it.
I am curious by nature. I read, I listen to podcasts, I save articles. But when I actually needed that one useful framework from a podcast six months ago, it was buried across notes, tabs, and saved links.
inding it took longer than starting over. And even when I found the source, asking a general chatbot about it gave me a confident answer with no way to verify it was real.
NotebookLM fixes exactly this. You upload your own sources, PDFs, YouTube videos, articles, research papers, your notes, and it only answers from what you gave it.
It does not invent a statistic that sounds reasonable. It stays inside your material and cites exactly where each answer came from, so you can check everything.
Then it goes further. It can turn your research into a quiz, a structured summary, a mind map, even a podcast with two AI hosts discussing your notes.
I use that on long research sessions, and it is one of the best ways I have found to actually retain information instead of just consuming it.
When accuracy matters more than speed, this is the tool.
6. [Granola](https://www.granola.ai/) (meetings without the bot)
Let us talk about meetings. Most of us spend a huge chunk of the week in them, half listening while half scribbling notes, and end up with neither good notes nor a clear memory of what was decided.
Most meeting tools try to fix this by sending a bot into your call. That is my biggest pet peeve. You have been in the meeting where an uninvited bot joins and everyone gets annoyed.
Granola works differently. It uses your system audio to capture the meeting in the background. No bot. Nobody knows it is there.
You can jot rough notes if you want, and afterward Granola merges your notes with the full transcript into something structured and actually useful.
The use cases stack up fast. A salesperson can turn the transcript into a follow up email with every detail from the call already in it. And the best part is what it gives back in the room. You get to be present in the conversation again.
Pro tip. Connect Granola to your everyday AI assistant through a connector, and at the end of a meeting it can push the next steps into a task manager like Notion or Asana without you lifting a finger.
7. [Google Nano Banana Pro](https://gemini.google/overview/image-generation/) (visuals in plain English)
This last one is for anyone working with images. It cuts your production cost and time hard.
Some of you are wondering why not Midjourney. Midjourney is amazing if you are technical. But for most people, Nano Banana Pro is far easier to start with.
I use it for content visuals that go alongside my videos and posts, infographics, whiteboards, and images for LinkedIn.
What makes it different is precise editing in plain English. No special syntax, no hours tweaking parameters. You describe what you want and it delivers.
Not right? Iterate. Say remove the background, change the color, make it more minimal, and it adjusts. The result comes out photorealistic.
Pro tip. If you need a 4K high resolution image, you can access Nano Banana Pro through Google AI Studio and buy extra credits to download at the settings you want.
The whole stack at a glance
Here is the complete list, from chat to visuals:
Finally…
If I had to start from zero today, this is the stack. It is not a lot, and that is the whole point.
The people who get ahead with AI are not the ones using the most tools. They are the ones who pick the right few, learn them deeply, and build a personal system around them that actually saves time.
So do not try to do it all. Pick one tool from this list today, the one that solves your most urgent problem right now.
Use it for a real task this week. Then add the next one. That is how you build something that sticks.
Stay curious, talk to you tomorrow.