FAQ Frequently asked questions

“What happens in a call?”

It’s like having an experienced developer looking over your shoulder. You share your screen and show us what’s happening. We’ll try to help by asking you questions, giving you suggestions, and walking you through manual debugging, etc.

“How quickly can you help?”

Sometimes within minutes, but usually the same day. Check our online scheduler to see when the next time slot is available.

“Can I get a refund if I’m unhappy?”

Absolutely. If you are dissatisfied with our help we will refund your fee no questions asked.

“What if I’m not ready to commit?”

No problem. Bookmark our website and keep us in mind for when you really need some help. You can also reach out by chat or email to ask any questions before you commit to scheduling a call.

Remember that the real cost of using AI is your time. If you are spending hours on something and not making progress it might be time to give us a call.

TechStuff we use most

“What AI tools do you recommend?”

• GitHub Copilot is a great all-rounder and their subscription prices are hard to beat. It integrates into VS Code or has a CLI available if you prefer working in a terminal. It supports multiple models across several vendors and has pro access to them built into the GitHub subscription price.

• Cursor is great if you don’t mind paying more and using a special version of the VS Code editor. While it was our recommendation earlier this year, GitHub CoPilot overall is a better value proposition.

• Claude Code is a CLI that works directly with Anthropic’s Claude models. It’s simple, effective, and a good option if you’re already subscribing to Claude directly.

“What AI models do you recommend?”

• Anthropic Claude Sonnet 4.5 is an amazing AI model for software development. It’s our clear favorite.

• We just started experimenting with Claude Haiku 4.5 which is billed at 88% of Sonnet 4.5 and 1/3 the cost. It’s promising so far.

“What programming languages do you recommend for AI-assisted coding?”

• We recommend Go (aka golang) if you’re able to choose a language. It has a very “picky” compiler that enforces consistent coding habits, has a massive standard library, a growing 3rd party ecosystem, compiles to a single executable that deploys anywhere, and works well with a strong coding AI like Claude.

• Python and JavaScript are good for humans but more difficult for AI models. The expressive freedoms these “relaxed” languages provide can often slow down development with AI agents and produce messy code that’s hard to debug.