Lumi | Devpost

Inspiration

According to Nieman Labs, around 40% of Gen Z prefers TikTok and Instagram over Google for getting their information. As the amount and forms of data available to us increase, it can also become more difficult to sort through all of this information and easily find what we're looking for. So, we created Lumi, a search engine that does the heavy lifting and provides users with the most relevant information that exists across various platforms.

What it does

Lumi takes in the user's search and consults a variety of sources, including Tiktok, Reddit, and Chat GPT-4. Our search engine ranks these results in order of relevance and provides users with an easily understandable combination of the most useful information from the most popular platforms!

How we built it

We wanted to create a search engine with a scope beyond Google, so we accessed the APIs of several sources ranging from Tiktok to IMBd. We used SvelteKit and VSCode to handle the full stack nature of the application. The SQL databases were stored in Google Cloud. For the front-end, we used typescript and tailwind css to ensure a visually appealing look for users.

Challenges we ran into

Initially, we struggled to create a tiling algorithm that would present information from several different sources in a way that was still quick and easy for users to consume. We also had issues with the lack of accessibility with APIs, as many of them had restrictions on their use and were tedious to blend seamlessly into our project.

Accomplishments that we're proud of

We were proud of our use of APIs, as we initially struggled to effectively use them. There were restrictions on us accessing some of the the most used APIs associated with our target sources, so we had to figure out the procedure for using APIs that were less known and less conducive to our exact goals. However, we were still able to get the results we needed from the APIs and present them in a way that we're proud of! We were also proud of our planning process! Lumi is something that all of us would use in our daily lives and a more ambitious project than we were used to, so we are happy with the results!

What we learned

None of us had used SvelteKit or Tailwind CSS before, so building Lumi with these tools in one weekend gave us a quick yet thorough look into their features and benefits. We also used a variety of APIs that each had different procedures, which provided us with a useful knowledge about API use.