An Uzbek AI assistant prototype in ten minutes
A user's prompt, one plan, 26 credits, ten minutes. What came out, what is real, what is honestly a placeholder, and why the prompt worked.
Ilyos Olimov · · 4 min read
Most prompts people type into AppX are one sentence. "Make a fitness app." "A shop for my bakery." Those work, but the plan has to guess a lot.
This one did not guess. A new user wrote it on his phone on a Sunday evening, and it is close to a perfect brief. I am sharing it, with the app name changed and his details left out, because it shows what "describe your app" can mean when you take it seriously.
The prompt
Create a mobile-first AI assistant app called Yordam. The app must have an original futuristic design, with a dark background, a glowing blue animated AI core in the center, and a clean premium interface.
Main screen: Yordam logo and name. A central animated AI orb. A microphone button for voice interaction. A text input field for messages. A conversation area where the assistant's responses appear. Uzbek language as the primary language. A settings button. A memory section for saved notes and goals.
Make the interface responsive for iPhone screens.
Start with a working prototype. Do not pretend that AI voice, long-term memory, or autonomous actions work unless they are actually implemented. Structure the project so real AI services can be connected later.
Keep the design original. Do not copy another company's logo or interface.
Notice the last two paragraphs. He is not asking for magic. He is asking for an honest prototype with a clear place to plug the real thing in later. That is exactly the right ask.
I added one paragraph of build notes: all visible text in Uzbek Latin script, the chat should work offline with a local assistant that gives simple Uzbek replies, notes and goals saved on the phone, and the microphone button should say "voice mode coming soon" instead of pretending to listen.
What happened
| Step | Time | Credits |
|---|---|---|
| Understanding the idea, planning the screens, designing the look | 1 min 34 s | 5 |
| Review the plan, one milestone, estimated 27 credits | my click | |
| Building the screens, 13 files | 8 min | 21 |
Total: just under ten minutes and 26 credits. The free account comes with 100.
What came out
Three tabs: Suhbat (chat), Xotira (memory), Sozlamalar (settings). I am describing the screens instead of showing them, since the app is his.
Chat. A dark screen, the app mark, an animated blue orb with orbit rings, and a small badge that says "Offline prototype, local replies". The conversation works. You type "Help me plan my day", it answers in Uzbek with a sensible suggestion. It is a local rule-based assistant, and the app says so on the screen, which is what the prompt asked for.
Memory. Reminders and goals, saved on the phone. Two sample reminders, two sample goals, a filter row, edit and delete. The header says "Saved on your phone".
Settings. This is the part I liked most. It has a section called Services with two rows: "Connect to a real AI service" (the local assistant is active now, a server address and key can be entered later) and "Voice mode" marked "coming soon", with a plain sentence that the microphone currently only shows a notice. The app tells the truth about itself.
The microphone button does exactly that: tap it and it says voice mode is coming soon.
Why the prompt worked
- It names the screens. The plan did not have to invent them.
- It names the language. Half of the prompts I see leave this out and get English.
- It states what must not be faked. The builder followed it and put the honesty in the UI, not just in the code.
- It describes a look in two adjectives and three colours, not in a paragraph. Enough to design, not enough to argue about.
What I would do next
Connect a real assistant API in Settings, which is a ten-line edit by chat. Add a daily summary screen. Then a milestone for push reminders. None of that was in the prompt, so none of it was built, which is also correct.
If you want the same thing for your own idea: write the screens, the language, and what must stay honest. The rest the plan can handle.
Keep going
More practical ways to move your app forward.
- How to describe your app to AI: a prompt template
Write a clear AI app-builder prompt with a copyable template, a practical example, acceptance checks, and focused follow-up edits.
- How to turn a client brief into a mobile app prototype
Turn a client brief into a mobile app prototype with a copyable AppX prompt, a practical test checklist, and a focused client-feedback workflow.
- Do not pretend the feature works
A user wrote one line in his prompt that every prototype should carry: do not pretend that AI voice, memory or autonomous actions work unless they are actually implemented. Why an honest prototype is worth more than an impressive one.
- You can build your app with a coding agent. Here is why most founders should not.
Coding agents in the terminal are the best thing that happened to developers this decade. For a founder with no developer, they are a maze with no map. What the first week looks like with an agent, what it looks like with AppX, and when you should still pick the agent.
- Three emails a solo founder sends, and what came back
AppX now sends three short emails from me: when your first app is ready, when you built one and did not come back, and when you signed up and never started. 56 sent so far. Here is what they say and the mistake that means I cannot tell what came back.
Ready to try your idea?
Start with a clear description. AppX will help you turn it into an app you can test.