NameWiz: Building a Creative Name Generator with GitHub Copilot

#devchallenge #githubchallenge #webdev #ai

This is a submission for the GitHub Copilot Challenge : New Beginnings

What I Built

Meet NameWiz—your brainstorming buddy for coming up with creative names for your next big idea. Whether you’re launching a tech startup, opening a cozy café, or starting a blog, this app makes the naming process easy.

Just type in a few keywords, choose a category, pick a tone, and even select a language. NameWiz will do the rest, giving you ideas to kickstart your project.

Demo

You can try the app here: NameWiz on Streamlit

If you’d like a quick preview before diving in, check out the video below. It shows how easy it is to use the app—just type in some keywords, pick a category and tone, and watch the names come to life.

Repo

The code for NameWiz is available on GitHub:
GitHub Repository

Copilot Experience

I began NameWiz by asking GitHub Copilot to create a basic Streamlit app and set up the project structure. Here’s the first prompt I used:

The output wasn’t perfect, but it gave me a clear direction to start building.

While GitHub Copilot was helpful, there were a few things that could be better:

Chat Limitations: Clicking "Done" in the chat clears all history, which made it tricky to go back and see what worked before—I couldn't find the history.
Broken Markdown: Some of the markdown it generated needed a lot of manual fixing.
Inline Chat: The inline editor was good for small tweaks and quick fixes but struggled with more complex tasks, like writing a proper README.md.

I mostly used the chat to get help with tweaking the UI, figuring out prompts for the model, and coming up with ideas to improve the app. It was really useful for brainstorming and fine-tuning things, even if it didn’t always get everything right.

GitHub Models

For NameWiz, I used GitHub's GPT-4 model to generate creative business names. The model was integrated into the app using a GitHub token for authentication. Since the app relies on this token, it might encounter rate limiter issues during heavy usage.

Here’s how I structured the prompt to interact with the model:

System Message:

"You are a creative business name generator. Only return a list of names without any descriptions or additional context."
User Message:

"Generate {count} unique and {tone} business names based on the combined keywords '{combined_keywords}' in the '{category}' category. Context: {context}."

The app also allowed users to generate names in different languages by incorporating the selected language into the context:

"{context} Generate names in {language} ({selected_language})."

By passing in the keywords, category, tone, and language, the model created names that reflected the user’s preferences. For example, entering keywords like "eco" and "travel" with a fun tone generated playful and relevant names.

Conclusion

Building NameWiz was a great way to explore how AI can help with creativity. GitHub Copilot and GPT-4 models made the process faster and gave me ideas I wouldn’t have thought of on my own. While there were some challenges, it was exciting to see the app come to life and imagine how it could be useful for others.