People always ask me what AI tools I use. As someone who builds AI systems for a living, I've tried virtually everything. Here's what survived the hype and became part of my daily workflow.
Coding: Claude Code
This is the tool that changed how I write software. Claude Code isn't just autocomplete — it's a coding partner that understands your entire codebase.
What I use it for:
The key insight: Claude Code works best when you give it context. I always start by explaining the project structure and constraints before asking it to write code.
Research: Perplexity
I stopped using traditional search for technical research. Perplexity gives me direct answers with citations, which is exactly what I need when I'm debugging a deployment issue at 2am.
My workflow:
Communication: ChatGPT
For drafting emails, meeting summaries, and communication in general, ChatGPT is still my go-to. The voice mode is particularly useful for brainstorming while walking.
Design: Figma + AI
Figma's AI features have gotten surprisingly good. I use them for:
Writing: Claude
For long-form writing — blog posts, documentation, proposals — Claude produces the best output. I write outlines and let Claude expand them, then edit heavily. The result is my voice with AI-assisted structure and depth.
Knowledge Management: Notion AI
Notion AI helps me:
Terminal: Warp
Warp's AI command suggestions save me from constantly googling CLI syntax. Type what you want in natural language, get the exact command.
What I've Stopped Using
Not every AI tool sticks:
The 3-Hour Savings
Here's roughly how the time breaks down:
That's 3.5 hours per day. Over a year, that's over 900 hours — basically 6 months of extra productive time.
My Advice
Don't try to adopt every AI tool at once. Pick one workflow where you feel the most friction, find the best AI tool for it, use it for two weeks, and measure the difference. Then move to the next one.
The compound effect is real. Each tool you adopt makes you a little more productive, which gives you time to learn the next one.