24/3/2016

Datamatic

 

Datamatic allows you to create complex, interactive infographics and data visualization projects in an intuitive, easy-to-use interface. Since 2014, more than 15,000 users in 144 countries have tried it. We spoke to the project's Founder & Developer, Jaroslav Benc, to find out more.

DDJ: Can you tell me a little bit about your background and how you came to work on Datamatic?

JB: I am a software engineer primarily focused on building web-based tools for designers and developers. I started Datamatic as a side project back in 2014. The main trigger was that I was giving a talk where I needed to present hierarchical data. There were plenty of D3.js examples, however I couldn’t find an easy editor where I could quickly edit data, change colors, and so on. I had some spare time during the Easter holidays so I decided to run a personal five day hackathon and create a simple tool for my presentation.

What was the inspiration behind Datamatic?

Datamatic was a side project at the time, so the main driving force was my interest around D3.js and data visualizations generally. I really wanted to get hands on it and see what kind of stuff comes out when designers start “drawing with data.”

The first public prototype was based on D3.js and had some attention, but wasn’t too useful. I shared an article about this on Medium.

datamatic1.PNG

I decided to experiment and create a version based on Brian Foo’s Transit Map:

datamatic4.png

This was supposed to illustrate the idea of  editing pieces of code that people like Brian wrote, but didn’t have much success with this either.

Then, I found out that Baidu is working on a pretty amazing library, ECharts, that very few people knew about it at the time. So, I decided to run another experiment and integrate that.

datamaticscreen.PNG

This was finally something that people started using. Hundreds of charts were created and sites such as http://www.datavizcatalogue.com started pointing to the tool. That gave me a bit more confidence to work on this further.

Datamatic is built on Google technologies. Why were these chosen?

As a developer focused on building editors, I’m fascinated with how Google makes collaborative editing, such as with Google Docs, easy and intuitive. When I found out that Google provides the whole editing capability as an API service, I was really keen to give it a try. Another reason that really resonated with me was building an app on a completely scalable and serverless architecture. No servers, no virtual machines, just publicly available APIs!

What makes Datamatic different from other data visualization tools?

For now, the main difference is that companies can create their custom theme and use it automatically in any chart they create. For instance, the following example was created using a custom Perceptica Media Research theme:

dataper.PNG

When we started working with Perceptica, their designer created a theme which consists of a set of colors, images, font sizes, and so on. Now, when journalists create a chart, they simply pick a theme and don’t even bother with selecting individual colors at all. This maintains visual quality and consistency.

What data formats are compatible with Datamatic?

Datamatic currently only supports copying/pasting into the data editor (similar to Excel). Since it’s not a BI or data analytics solution, it simply offers a way to edit data visualization templates. Each template then defines what kind of data it requires. For instance, a simple pie chart is usually a list of names->values and treemap is a tree of names->values. Datamatic handles tree data using a breadcrumb on the top of the editor:

datamaticdata.PNG

Another option is to use API for each chart.

datamaticapi.PNG

This applies when you need to connect some external API endpoint, or if you want to connect your chart into external UI. Here is an example and tutorial:

datamaticlondon.PNG

Can you tell us about the databank that you are working on with Datová ┼żurnalistika?

Recently, we were selected as one of the Google Digital News Initiative winners. We’re going to work on an online marketplace where journalists can browse and buy ready-to-publish data visualizations or even whole data stories. The point is that all data visualizations made in Datamatic are easily editable, meaning that journalists across Europe can buy, translate, customize, and re-publish data visualizations in a matter of minutes, rather than hours or days. It’s a bit like an image bank but for data visualizations.

Another exciting direction is that Datamatic has an built-in code editor to develop new templates (Datamatic IDE). I’m hoping that this feature will easily connect data visualization developers and data journalists.

We’re very open to anyone who’d like to hear more and collaborate on this project. This may include data visualization developers (to create templates), data journalists (to create stories), or media organizations (to pilot/test this project). Feel free to write us at .(JavaScript must be enabled to view this email address).

Datamatic is partnered with Perceptica Media Research, SeeNews, and SME. Can you tell us about how you work with these partners?

First of all, I’m very thankful to these early adopters who provide us with invaluable feedback. This is one of the joys of running a startup project. Seeing your side-project becomes actually useful to someone is a very magical experience.

We’ve gone through a number challenges and learnings with each of these partners. For instance, SME was a great challenge to make the whole infrastructure scalable and as cheap as possible. SME is the most popular news site in Slovak Republic, so when they publish an article, we have to serve several hundred requests in a very short time. This means that we have to use serverless architectures, services such as Amazon S3 and so on. We also need to drive cost of hosting to as little as possible so we can offer affordable plans to our customers.

With SeeNews, we had to support custom themes per customer, which now seems to be a very useful feature. We also take advantage of Google APIs where we can easily share visualizations, collaborate in real-time, and so on. For instance, if a data visualization template doesn’t have a property that customer needs (color of something, formatter, etc.), we can open the file and add it in real-time. The customer doesn’t even need to refresh or do anything, it’s just Google real-time API magic (thanks Google!).

We’re also more than happy to hear what kind of templates customers want. They can now sponsor a template for a flat price of $20. This may include adding a new map, parliament templates, timelines, and so on.

Try Datamatic for yourself! Visit the website here.

Comments