Today, we’re joined by Gregory, or you may know him as GregoryCongad on Github, for yet again one of our Meilistars interviews!
Meet Greg!
Let’s start by learning a little bit more about our latest Star: just last spring, Greg finished his undergrad in computer science at WPI and started contributing to Meilisearch at the end of 2022.
He actually got introduced to Computer Science all the way back in 5th grade. That’s 10 years old for all non-Americans out there! And that’s all thanks to him stumbling across Scratch, the famous MIT website. It was love at first sight and Greg has been passionate about programming ever since.
But he didn’t stop there!
“In middle school, I self-taught Python, C++, and Java by reading some books and by coding up some miscellaneous side projects. I've dabbled in many languages and projects over the years, but have finally picked up Rust after hearing so many good things about it (and the hype is there for a reason!)”
Truly a young prodigy! And that’s not even half of it, for the past two summers, Greg has been interning at Amazon under the Alexa project, focusing on internal tooling, creating products for internal consumption across the organisation. Not much more can be shared on this matter of course due to his NDA, but this sure looks like high profile work.
A serendipitous discovery
With all this experience and knowledge under his belt, we couldn’t help but wonder how he’d discovered Meilisearch. And why change a successful method? Just as with Scratch, Greg confesses he simply stumbled upon Meilisearch:
“I was looking for a full text search engine to embed in a Flutter library and was going to settle on TypeSense. TypeSense wasn’t exactly what I wanted though, and eventually I found Meilisearch through one of TypeSense’s comparison pages in their docs. I realized Meilisearch was perfect for my use case, especially since I saw it was written in Rust and I wanted to pick up Rust at the time.”
We really got lucky to have caught his attention! We wondered what had convinced him to give Meilisearch a try as opposed to TypeSense, and Greg shared two main reasons:
“1. Persistence on disk. Since I was embedding it, having persistence across application open/close was pretty important
2.The license. Meilisearch is licensed under MIT, whereas TypeSense is GPL. GPL is pretty restrictive, so I try to stay away from it whenever possible”
Our curiosity got the best of us once again, and we couldn’t help being nosey by asking which features of Meilisearch he preferred, and which one could do with a bit more work:
“My favorite aspect is how well everything is architected—it’s super easy to embed Meilisearch in other Rust applications and only grab the features you need. Of course, the typo tolerance and relevance that are builtin/require zero configuration are probably my favorite aspects of the product! As far as what I’d like to see, I’d love to have the engine auto-update/have compatibility across versions. Currently, you need to make dumps between versions, but it’d be very nice for that process to be automated (and some work is being done toward that end, like updating how Settings are stored internally)”
A lot of food for thought for all, which we’re very grateful for since hearing our users’ feedback is the best way for us to stay relevant and truly offer you all the best possible search experience.
And what of Greg's personal projects?
Seeing Greg’s impressive experience despite his young age, we asked whether he’s had the chance to use Meilisearch in a professional context or for a personal project:
“I’ve used/embedded Meilisearch to create Mimir, which is a Flutter library that gives developers an on-device NoSQL database with typo-tolerant & relevant full text search. I originally created Mimir with an application of my own in mind, but I’ve since put that application on hold. Now I just work on Mimir and contribute back to Meilisearch as a hobby project, since I enjoy working on both! The Meilisearch team is really welcoming and nice, which is a great change when compared to other open source projects. Really easy to land your first contribution!”
Go give Mimir some love if you have the chance! And we’re very touched to hear Greg’s first experience with us was a positive one, that’s definitely something we aim for!
Embedding Meilisearch to Mimir was actually how Greg first noticed that there were some points of improvement he could contribute to in order to make Meilisearch better! His first contributions were actually created as a way to solve the issues he ran into when embedding Meilisearch to Mimir! That PR was even featured in one of our release blog posts. Clearly, Greg and Meili were a match made in heaven.
A big thank you to Greg for his time and his answers, we were delighted to be able to spend a moment with him!