Next meal decision stressing you out already?

Context
At Sozcode, we are all about transforming ideas into real world solutions—and doing it fast.

Recently, we challenged ourselves to solve a common problem: when a group of friends can’t agree on where to eat, what’s the quickest way to find the best nearby restaurant based on everyone’s specific needs? With a little help from ChatGPT and some smart coding, we built a one-day project that uses Google Maps APIs and ChatGPT API to analyse nearby food options in Singapore and recommend the perfect spot on Gradio interface.

The problem‍

We noticed how choosing a restaurant often goes from “Let’s grab lunch” to “Let’s scroll through endless reviews and compromise.” Our diagram sums up the back-and-forth headache that happens every time you try to please everyone in a group. It’s tedious, it’s time-consuming, and more often than not, you end up settling for the same old spot when hunger gets the better of the group.

Turning ideas into action
‍Once we knew what our solution needed to do—collate everyone’s preferences, fetch restaurant data, summarise reviews, and then pick the best option—we asked ChatGPT for guidance on how to build it without wrestling with complicated local setups. We needed something that would be quick to prototype, accessible to anyone with a browser, and powerful enough to handle both data gathering and AI-driven summarisation.

Surprisingly, ChatGPT proposed an approach that ticks all those boxes. We ended up with an online environment for running code, a straightforward front end for user interaction, and two powerful services for pulling and summarising information. Best of all, it meant no local software installation or advanced coding experience was necessary—just a bit of curiosity and a willingness to learn.

Here’s the exact prompt we used to kick off ChatGPT’s suggestions (notice how we kept it broad, so ChatGPT could recommend the tools without us needing to know them by name):

"I want to create a simple online tool that helps a group of friends decide where to eat. It should collect everyone’s preferences, automatically look up local restaurants, summarise reviews, and recommend the best place. I'm not very experienced with coding, and I'd like to work in a free environment that doesn't require installations. What's a streamlined approach using easy-to-learn techniques and readily available tools for both the user interface and the AI aspects, focusing on rapid development?"
‍‍
Tools suggested

With the right tools in mind, an online coding environment, the Google Maps API, and the ChatGPT API, the next step is straightforward: ask ChatGPT to generate the full code for the project.
‍
Whenever we hit an error, we copied the error message and paste it back into ChatGPT. It can often pinpoint the problem and suggest a fix on the spot. This back-and-forth troubleshooting dramatically speeds up the development process, even if one have minimal coding experience.

If you are interested, you can review the comments left by ChatGPT to learn the function of every line and ask it to provide more examples to elaborate.
‍
Ready to give it a try?
Scroll down to see our final code snippet "What to eat.ipynb". (Viewable on desktop)
‍
You can also access our Google Colab file - just make a copy to your own Google Drive and replace our placeholders with your own Google Maps API and ChatGPT API keys for it to work!

Or if you just want to see it in action, check out our Instagram reels on this!‍ :D