← Back to all posts

Using MongoDB as a Graph Database

31 October 2014

Written by 


Chris Clarke


This is a presentation I gave to the NoSQL and Big Data Birmingham meetup group a couple of weeks back, and actually it’s a redux of a presentation I gave a couple of years ago at MongoDB UK.

It tells the story of what has been quite a journey for us over the last couple of years, as we migrated away from a general purpose graph database technology to managing our graph data in a document database.


The history here is that Talis used to be a graph database vendor of sorts; we had a PaaS offering called the Talis Platform, the genesis of which dates back to 2006. The TL;DR of that particular chapter of Talis history is that we decided in 2012 that our future was not in triple stores, so the platform was mothballed and access for external customers wound down.

However, the platform, and graph data, formed the architectural cornerstone of our growing EdTech business and its products. With the platform now EoL, we had some decisions to make about how to move forward with these products.

The presentation charts how we migrated to a new architecture based around document databases, which we call Tripod. Our choice of database happened to be Mongo but actually any number of document databases today would be an equally good choice.

You can review our work on Tripod via our github repos for PHP and for node, we’ll be blogging more about it over the coming months.

Meanwhile the slideshare presentation is embedded below. Be sure to check the presentation on the main slideshare site for the notes - most of the detail is in there :-)

Comments on HN