Start a project

How we use AI, and where we never cut corners

We use AI as a tool in skilled hands: drafts and variations go faster, while people stay in charge of strategy, code, facts, privacy and quality.

Clients now ask us, sometimes hopefully and sometimes warily, whether we use AI. The short answer is yes, every day, in design, development, marketing and production. The longer answer is the one that matters: how we use it, and the places where we never let it decide.

We are a professional studio, and we do things the proper way. AI helps us do them faster. It does not do them for us.

A tool in skilled hands

Almost every new tool has been announced as the end of the people who use it. Desktop publishing did not replace designers, and frameworks did not replace developers. They raised the floor, and the people who knew the craft went further with them. AI is the same kind of change, only quicker.

A model can produce a paragraph, a layout or a function in seconds. It cannot tell whether the paragraph is true, whether the layout suits the brand, or whether the function is safe to run against a client's data. Those judgments take experience, and they are what clients pay a studio for. So we treat AI the way a carpenter treats a power tool: useful, fast, and never left running unattended.

Where AI helps us

The best uses are the ones that save time on the first step, so that people can spend more of it on the steps that matter.

Designers and developers decide; the tools help them go faster
Designers and developers decide; the tools help them go faster

In practice, that means:

  • First drafts of copy, which a writer then edits, cuts and rewrites in the brand's voice.
  • More design directions to choose from early on, so the conversation with a client starts with options rather than a blank page.
  • Code suggestions while a developer works, read and judged by that developer like any other code.
  • Transcripts and subtitles for video, checked against the recording before they are published.
  • Summaries of research, as a starting point for reading the sources, not a replacement for them.
  • Faster picture variations for a campaign, once the direction and the brand are settled.

In each case, a person starts the work and a person finishes it. AI shortens the middle.

Where people stay in charge

Some parts of the work we never hand over, whatever the tool can do.

  • Understanding the client. That is the job: listening, asking the second question and noticing what a brief leaves out.
  • Strategy. What to build, what to say and to whom is decided by people who know the market.
  • Final design decisions. A designer chooses, and can explain why.
  • Every line of code that ships. Wherever it came from, a developer reviews it and it is tested before it reaches a client's users.
  • Brand guidelines. The rules come from the brand, and we apply them. A model does not get to reinvent them.
  • Facts and claims. Every figure, name, price and promise is checked against a source. A confident sentence is not a true one.
  • Arabic copy. It is written by native speakers, not translated by a machine and left as it is.
  • Data privacy. Client data is never pasted into public AI tools.
  • Quality checks. Someone who did not make the work checks it before anything goes out.

None of this is new. It is how good studios worked before AI, and it is what keeps the speed AI brings from turning into mistakes.

Working inside the brand, not guessing it

The most common weakness of AI-made pictures and copy is that they look like everyone else's. A general model knows what brands in general look like. It does not know yours, so it guesses, and the guesses drift further from the brand with every result.

Sketches and choices before anything is generated
Sketches and choices before anything is generated

Our sister companies build tools to close that gap. Melon Cut keeps a brand's guidelines as living pages: the logo, colours, fonts and voice, always current, at one address. Cutlin makes pictures and video that follow those guidelines, so the brand goes into every result instead of being added, or forgotten, afterwards. The AI then works inside the brand rather than guessing it.

Even then, the starting point is a person: a sketch, a direction, a choice between options. The tools come after the thinking, not instead of it.

The standard we hold ourselves to

The same team has worked on campaigns for brands such as stc and American Express, where every file answers to brand guidelines, legal review and a fixed launch date. Those are the standards we bring to every project, and AI does not lower them. If anything, it makes them more necessary. When producing a draft becomes easy, checking it becomes the real work.

Being open about it is part of the same standard. If you want to know where AI was used on your project, ask us, and we will explain where and how.

How Angry Monkey Agency can help

Angry Monkey Agency builds websites, platforms and apps, and designs and markets brands, with experienced people deciding and AI helping where it saves time without costing quality. If you want the speed these tools bring without the shortcuts that often come with them, we would be glad to talk.

Read how we work in software development, look at the work or start a project.

Chat with us on WhatsApp