5 AI Tools I Use Every Week to Save 10+ Hours (Real Workflows, Not Hype)

The AI Hype Problem

Everyone is talking about AI. Most of it is noise. "Use ChatGPT to 10x your productivity" is not a strategy — it's a tweet.

What actually works is specific. So I'm going to be specific. Here are the five AI tools I use every single week, exactly how I use them, and how much time each one saves me.

5 AI Tools I Actually Use Every Week

1. Claude — Writing, Research & Thinking Partner

Time saved: ~4 hours/week

Claude is my primary writing tool. Every newsletter issue, every long-form review, every email sequence — I start with Claude before I start typing.

My workflow: I dump raw notes, research, and a rough outline into Claude and ask it to help me structure a first draft. I then rewrite heavily in my own voice. The output is never published as-is, but having a solid draft to react to cuts my writing time in half.

I also use Claude for research synthesis. Instead of reading fifteen articles on a topic, I'll feed Claude the key pieces and ask it to identify the most important points and contradictions.

Prompt I use every week:

"Here are my raw notes on [topic]. Help me structure a first draft for a 1,500-word article aimed at [audience]. Lead with the most practical insight."

2. Perplexity AI — Real-Time Research

Time saved: ~2 hours/week

For research questions where I need current, sourced information fast, Perplexity has replaced Google for me almost entirely.

The difference: Perplexity gives you synthesised answers with citations, not a list of links to click through. For competitive research, pricing comparisons, and staying current on software updates, it's dramatically faster.

How I use it: Every time I'm writing a tool review, I start with a Perplexity search to get a current snapshot of pricing, recent changes, and what people are saying. Takes 2 minutes instead of 20.

3. Zapier AI — Workflow Automation

Time saved: ~2 hours/week

I use Zapier to automate the repetitive parts of my workflow. The AI layer lets me describe what I want in plain English and have it build the automation.

Best use case: Auto-tagging new email subscribers based on which lead magnet they used, then routing them into different email sequences automatically. Set up once, runs forever.

4. Otter.ai — Meeting Notes

Time saved: ~1 hour/week

I don't take notes in meetings anymore. Otter records, transcribes, and summarises every call. After the meeting I get a summary with action items highlighted.

Tip: Paste Otter's summary into Claude afterwards and ask it to extract all action items and decisions into a clean list. This takes 30 seconds and saves 10 minutes of review.

5. Fathom — Call Intelligence

Time saved: ~1 hour/week

Similar to Otter but built specifically for Zoom calls. Fathom records your calls, highlights key moments (objections, questions, commitments), and gives you a searchable library of every conversation.

For sales calls or client check-ins, this is invaluable. I've caught commitments I'd forgotten I made and spotted recurring objections I needed to address in my content.

What I Don't Use AI For

Being honest: there are things I've tried to automate with AI that didn't work.

• Final editing — AI edits make writing sound generic
• Strategic decisions — AI is good at synthesising information, not judgment calls
• Personalised outreach at scale — AI-personalised cold emails are obvious and don't work

The 10-Hour Breakdown

Getting Started

Start with one tool. Pick the one that fits your biggest time drain — writing or research — and use it consistently for two weeks before adding anything else.

I cover AI tools, productivity apps, and software every week in The Capital Stack. Subscribe free.