The 7 AI Tools I Actually Use While Building

Every week there’s a new AI tool claiming to 10x your productivity. Most of them are noise.

Here are the 7 that actually made it into my daily workflow — and why.

1. Cursor IDE

Stop using VS Code with the GitHub Copilot extension. Cursor is built around LLMs from the ground up. The difference in experience is significant.

What makes it stick: the Composer feature lets you describe a change in English and it edits multiple files at once. For refactoring tasks, this alone saves hours.

2. Claude (Anthropic)

For complex reasoning tasks — architectural decisions, debugging gnarly logic, writing technical docs — Claude consistently outperforms the alternatives for my use case.

The long context window matters. You can paste an entire codebase and ask questions about it.

3. OpenRouter

One API key, every model. When Claude is at capacity, I fall back to Gemini or GPT-4. OpenRouter handles the routing. Stop managing 10 different keys.

4. Perplexity

Google for builders. When I need current information — not just what the training data knew — Perplexity gives me cited, up-to-date answers. I use it for market research and competitor analysis constantly.

5. Notion AI

My second brain. Every project, every build log, every random idea goes into Notion. The AI features let me summarize meeting notes and generate first drafts from bullet points.

6. Make (formerly Integromat)

If you’re still writing custom webhooks for everything, stop. Make lets you chain together 1,000+ services visually. I use it for automating repetitive data flows between APIs.

7. ElevenLabs

Text-to-speech for content repurposing. Write once, get audio. Not essential for everyone — but if you’re building any kind of media or content pipeline, it unlocks fast.

---

The pattern you’ll notice: none of these replace thinking. They compress the time between idea and shipped. That’s the metric that matters.

What tools are in your stack? Reply to this on X: @cuccicoffee