GitLab is banking on attracting 1,000 users per month for its Meltano platform by the end of this year, as it becomes the latest firm to try and bring a bit of collaboration and version control to the data space.
The CI/CD/Dev(sec)Ops vendor first unveiled the platform in August 2018, before going uncharacteristically quiet. However, at its London shindig earlier this month, it announced that Meltano had officially hit v1.0, and could be installed/run from Digital Ocean.
The project is now headed up by Danielle Morrill, whose previous credits include being founder and CEO of Mattermark.
Speaking to DevClass earlier this month, Morrill said the scope of the project had shifted over the course of the year from being predominantly focused on developers, to encompass data scientists.
“If you want to go into the data world, you have to broker the relationship between the data scientist, data engineer and the data analyst. And that’s a huge range of technical skill,” she said. “I mean, the scientist is a mathematician. The analysts literally could be like freshly minted from Business School. It’s a huge range.”
So, she continues, the aim is to bring the GitLab approach to the software lifecycle and apply it to the data ops world. “They have different problems, because they have different tools, but it has the same tool chain chaos, basically.”
“So v1 is really about deployability to the cloud, so that you don’t have to do any of that hosting work. So you can go from nothing to a dashboard, in 10 minutes or less. And really just delivering on the promise of having that workflow.”
Doing so would massively increase the potential audience for GitLab, she said. “It makes it accessible to the 10 million people in the world who are data analysts,…potentially doubling the market opportunity for GitLab. These people are almost software developers, but they don’t get treated like software developers”.
Morrill accepted there was some scepticism about GitLab trying to enter the data market. But, she countered, this was changing, as evidenced by contributions to the project.
“In the past six months, we’ve really gone from, ‘okay, we’re thinking about this’, to ‘oh, we can really see it coming to fruition’. And part of the way you see it is when you start getting external contributions. Other people are like, we see where you’re going. “
Presumably, Morrill would like those contributors to help it fill some of the project’s admitted gaps. As she put it, the project chains together existing open source tools – currently Stitch, DBT, Apache Air Flow and Jupyter Notebook. While the Meltano team has developed modelling and data visualisation components, it still hopes to identify open source equivalents to fill those gaps.
Looking further, she said, “The big thing going forward is version control. We don’t have that yet.”
Further still, the aim is to extend the platform’s collaboration potential – an area where data science falls down, with most data scientists working on their laptops, rather than collaboratively, across the cloud.
“So right now people mostly use Montano by themselves. Most data people still do their work on a laptop. So it’s in their local memory. It’s like if you have a laptop on an aeroplane, too bad, like you can’t collaborate. So that’s a way in which the data ops world is really behind.”
And who is Meltano itself behind? Asked who the competition for the platform is, Morrill reels off Tableau, Periscope, and Looker. It’s an ambitious hit list for the six people strong Meltano team at GitLab – Tableau sold out to Salesforce this summer for $15bn.
While those tools are thought of as being aimed at virtualisation, Morrill argues, “I think the companies are buying the same value prop, which is like you’re going to pull your data out of somewhere, you’re going to put it here and you’re going to understand what it means.”
As for GitLab, “We’re really behind in the sense like I’m not going to be able to build Tableau in six months….what we can do is be ahead of them on being fast with the end to end experience, because it’s going to be tough for Tableau to do that.”
For GitLab, she says, data science represents “another $1bn opportunity”. But there’s a broader angle, Morrill argues.
“I think the goal would be GitLab figures out a recipe for building new vertical end to end applications. There’s other workflows that could be automated, like marketing automation, for example, is a huge stack.”
As GitLab prepares to go public, she says, “We need to tell a story not just about our existing products in our existing vision, but our ability to continuously innovate on all the areas. I would love to see us, build a $50bn company just on what GitLab is today and continuing. But, you know, you can grow an even bigger business by having these other verticals.”
In the meantime though, Meltano has to build a user base. Right now, Morrill says, it’s at around 300, but “Our goal is to be at 1000 monthly active users by the end of the year. So you can hold me to that… And if not, we can share a little bit more about what we’ve learned.”