The tools I actually use.
Every week I work across dozens of projects. That means I spend real time inside these tools, not just reading about them. If something on this list stops being useful, it gets replaced. If something new works better, it gets added.
This is not a list of every AI product that exists. It is what I reach for when I am solving real problems for real clients. Code editors, LLMs, APIs, vector stores, MCP tooling, and the development stack I build on top of.
Most of my client projects are built with Next.js, TypeScript, and Tailwind, deployed on Vercel or similar platforms. The AI layer changes depending on the project. Sometimes it is a chatbot powered by the OpenAI API. Sometimes it is Cursor generating an entire feature branch. Sometimes it is Claude reviewing a pull request.
The point is not to use every tool. The point is to know which one fits each situation. That is a big part of the work I do with clients: cutting through the noise and picking the right tool for their specific problem.