If not done right from the beginning, data governance can be a huge pain for data teams.

João Maia, Head of Data at Quanto, can attest to this from years of experience. Having previously worked on teams where the focus was on delivering results without worrying about data governance, João has seen what happens when you do that:

  • Everything starts off just fine. Then the company starts growing.
  • New people and processes are introduced into the mix.
  • Data gets scattered around different databases and tools.
  • Losing track of information becomes treacherously easy.

Sprinkle on a little bit of everyone’s favorite data regulation laws like GDPR and LGPD (Brazil’s General Data Protection Law), and the scenario quickly becomes chaotic. And once you’re past that point… Well, we hope your hobbies include herding cats and reversing entropy.

That’s why, when João was presented with the challenge of being the Head of Data at Quanto, he decided that things would be different this time.

How not to make the same mistake twice

In operation since 2020, Quanto is a Brazilian Open Banking/Open Finance company transforming the potential of Open Finance into business results. In a secure environment, customers can share data and extract insights to better understand the consumer profile and build more personalized financial solutions.

Fintech isn’t exactly known for a happy-go-lucky, relaxed attitude towards data, so João was determined to get Quanto’s approach right from the start.

That included identifying the main data-related mistakes he’d seen in the past:

  • Decentralized data. When data gets scattered around too many different places, it’s easy to lose track of what information is useful, and to who. But for data to be actually meaningful, it needs to be clear who is responsible for it.
  • Lack of visibility into impact. There must be a better way to understand the impact of data than just deleting a column and waiting to hear which desk the scream is coming from. When you can’t predict the impact of moving data from point A to point B, you’ll spend a lot of time trying to figure out if it can be done without breaking anything. This is time that most companies simply don’t have.
  • Lack of visibility into data usage. In terms of usefulness, not all data is created equal. If your engineering team can’t see what is important and what isn’t, they have no way to prioritize what and how they’re monitoring.
  • Outdated and unused trash data. Data can become outdated and tables unused over time, for different reasons. Holding on to unused data causes clutter and confusion for your engineers and data scientists, not to mention the precious, precious server space it’s taking up.

That’s quite a hefty list of exciting ways things can go wrong. João’s first impulse was to develop an in-house tool to deal with these issues before they even come up, but after some consideration, he realized that wasn’t feasible for Quanto.

Then, João and his team tried an open-source tool, only to find out it didn’t have the level of automation they needed. Too much manual work = an effort-to-result ratio that wasn’t making sense.

Next, they looked into three different data governance products. Moving into the proof-of-concept phase for each of these tools, it quickly became obvious that they would take an unreasonable amount of work to configure and integrate properly.

And then they met Alvin.

It was a LinkedIn post from Lucas, one of Alvin’s developers, that planted the idea in João’s mind. He was intrigued. After a demo and a proof of concept, it turned out to be a perfect fit.

How Quanto uses Alvin

“We plugged our platforms into Alvin and it just worked: lineage, impact analysis, and glossary”, says João. “No more time spent configuring things manually and checking every day if they’re working.”

We had to dig a little deeper, so we asked João for specific ways in which Alvin makes his life easier.

“We chose Alvin because of how easily, quickly, and efficiently we could implement the LGDP controls,” he says. “Unlike other options in the market, whether open-source or proprietary solutions, that take weeks to deploy, with Alvin, we did it in minutes.”

"I always like to use the “Dashboards” filter to understand what the data team’s internal customers are consuming, to help prioritize the team’s efforts."

“When visualizing lineage, I can see which Looker Explores are being used to build a dashboard and which tables are used. This type of investigation would be more work with ETLs or on Looker itself. With Alvin, it’s much simpler and in one place.”

“The impact analysis feature is also very useful to understand the implications of changing a table that is already part of some productive flow of value delivery.”

Sounds great, we’ll take it!

Wait… We built it.

And we couldn’t be prouder to see how Alvin is helping Quanto knock it out of the park with data governance, centralizing data information, and democratizing access to it.