AWS Builder Center
Weekend Showcase Challenge: WishWeave

Weekend Showcase Challenge: WishWeave

I built WishWeave from a simple AWS Builder Center Wishlist idea to help people turn rough project or feature ideas into clearer, more actionable starting points.

Vision and what it does

The idea for WishWeave started with a small problem I noticed while using the AWS Builder Center Wishlist.
Earlier this summer, I submitted a wishlist idea called “ Show Related Wishes Before I Post .” My thought was simple: when someone is about to submit a new wish, it would be useful to see similar wishes first. That could help people find an existing discussion instead of creating another wish about almost the same thing.
That idea became the starting point for WishWeave.
WishWeave is a small tool for turning a rough project, feature, or wishlist idea into something easier to think about. Instead of trying to write a perfect description at the beginning, you can just write what you have in mind. WishWeave then turns that rough idea into a simple build canvas.
The canvas has four main parts:
  • The Problem
  • Who It Helps
  • Possible Solution
  • Next Small Step
It also gives some related ideas to help the user think about whether the idea may already have something similar, or whether there are other directions worth considering.
I wanted to keep the project simple. The goal was not to build a huge application with dozens of features. I wanted to make something I could actually finish and use.
The web version also includes a few example ideas and local saving, so users can keep their drafts in the browser while working on them. It does not require an account or an external API, and the saved drafts stay in the browser.

How I built it

I started with the original Wishlist idea and worked backwards from the question: what would actually help someone before they submit or start building an idea?
The first decision was to make the input very simple. A user should be able to write a rough thought in their own words without worrying about formatting it first.
From there, I decided that the output should not just be one large response. Breaking it into smaller sections makes it easier to read and gives the user a more practical starting point.
The four sections became the main build canvas. I also added the related-ideas section because that connects back to the original problem that inspired the project.
For the web version, I kept the implementation lightweight and self-contained. The project is available through GitHub Pages, and the source code is available in the public GitHub repository.
I then created a separate version of the same idea in PartyRock. I wanted to see how the same concept would work when the AI part was built using an AWS tool rather than the lightweight web implementation.
The PartyRock version follows the same basic idea. A user enters a rough project or wishlist idea, and the app generates the problem, who it helps, a possible solution direction, a next step, and related ideas.
I also made an important distinction in the app itself: the related ideas are AI-generated suggestions. They are not results taken from the actual AWS Builder Center Wishlist or another real database. I wanted that to be clear rather than making the application look like it had access to information that it does not actually have.
One challenge was making the related-ideas feature useful without giving the impression that it was searching the real Builder Center Wishlist. Since this version does not connect to a real wishlist database, I kept the suggestions clearly labeled as AI-generated and treated them as a way to help users think about their idea rather than as actual duplicate-wish results.

AWS services used and architecture overview

For the AWS version, I used PartyRock, an Amazon Bedrock Playground. PartyRock is a browser-based tool for building generative AI applications, powered by Amazon Bedrock.
The basic flow of my PartyRock version is:
Rough idea → AI-generated sections → related ideas
PartyRock provides the app interface and connects the different generation steps, while Amazon Bedrock provides the generative AI behind the responses.
The user's idea is the main input. The different parts of the app use that input to generate the problem, who it helps, a possible solution, the next small step, and related ideas.
I kept this version focused on the same purpose as the original WishWeave project rather than trying to add a completely different set of features.
The GitHub Pages version and the PartyRock version are not the exact same codebase. They are two versions of the same project idea. The GitHub Pages version is my lightweight web prototype, while the PartyRock version is the AWS-based generative AI version built using PartyRock and Amazon Bedrock.

What I learned across the summer

The biggest thing I learned from the Summer Build Series is that a project does not have to start as a big technical system.
My original idea was basically one sentence: show related wishes before someone posts a new wish.
Turning that small idea into something usable made me think more about the actual person using the application. What should they enter? What should they get back? What information would actually help them take the next step?
I also learned that keeping a project small can be a good decision. While building WishWeave, I thought of several features that could be added later, but I tried to keep the main purpose clear instead of adding things just to make the project look bigger.
Another useful lesson was learning to work with different tools. Building the web version and then building the same basic idea in PartyRock gave me a chance to see the difference between a simple web implementation and a generative AI application.
I am still learning software development and AWS, so I do not see WishWeave as a finished production system. For me, the important part was taking a real problem I noticed, turning it into a project, testing it, and getting it to a point where other people can actually try it.
The original wishlist idea could eventually become something more advanced, especially if it could use a real collection of existing wishes and better similarity matching. For now, WishWeave is my small attempt at exploring that idea and seeing where it can go.

Project links

A builder who inspired me

A special thank you to  Ben Fowler  and the AWS Builder Center team for organizing the Summer Build Series. Seeing the different projects throughout the summer gave me a lot of ideas and motivation to keep working on my own builds.
I especially liked that the challenges encouraged taking a simple idea and actually making something with it instead of leaving it as just an idea.
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article