MariaDB’s AI RAG Hackathon is underway

We had the pleasure on Friday to take individual calls with everybody who submitted to the ideation phase of the MariaDB AI RAG hackathon. 

The ideation phase deadline passed last week, and we are happy to share that we received several promising submissions for both the innovation track and the integration track. Innovation involves applications using MariaDB Vector, like RAG, and integration being enabling MariaDB Vector in an existing framework. 

Participants range from individual contributors to even a corporate team. Some already have some experience with AI, and some are newcomers to RAG. What they all have common is that they saw the hackathon as a chance to try out and learn something new. To scratch their own itch, so to say – and we’re thankful for that!

Inspired by the integration submissions, I put together MariaDB’s own list of existing and potential new Vector Integration Frameworks. You can probably guess that a couple of the integration submissions are included on the list. 

On the innovation side, people have pondered several use cases for using vector search including semantic search on Youtube content, analyzing massive user insight data and utilising graph database approach for more precise RAG results. 

We are really looking forward to seeing what the participants will develop by the end of the development phase on May 5th.The best submissions will get to demo at the Helsinki Python meetup on May 17th. 

Ps. In case you are pondering what to try RAG out on, one large data set we encourage trying is open data from Wikipedia and Wikidata. Wikimedia offers various data dumps, and even prevectorized wikipedia datasets can be found. An important aspect of RAG is also how you decide to chunkify the data before vectorising it. If you try it out, let us know what approaches you used!