Bigeye raises $17M Series A funding to boost data quality

The former Uber product manager and current CEO and co-founder of a startup outlines the challenges and opportunities of enabling a new data trust platform with data quality at its foundation.

Data quality startup Bigeye on Thursday said it raised $17 million in a Series A round of funding led by Sequoia Capital with participation from Costanoa Ventures.

Founded in 2019 as Toro and rebranded as Bigeye in November 2020, the company was started by engineers who had formerly worked at ride-share giant Uber.

Bigeye bills itself as a data monitoring platform with a focus on data quality, which is a key challenge for organizations dealing with large volumes of data.

Kyle Kirwan, co-founder and CEO of Bigeye, worked at Uber for five years, helping lead the development of Uber's internal data catalog known as Databook. Some of the lessons he learned from the experience inspired him and his co-founders to devise what they think is a better way for enterprises to build more trust in data to improve business outcomes.

In this Q&A, Kirwan explains what data quality challenges many organizations face these days and how Bigeye is building out its platform to help improve data.

Why did you start a data quality company?

Headshot of Bigeye CEO and co-founder Kyle KirwanKyle Kirwan

Kyle Kirwan: My co-founder [and CTO Egor Gryaznov)] and I were both early members on the data team at Uber. So he and I started Bigeye because a lot of the problems that we had to solve for the data engineering and data science teams internally at Uber, we realized exist for practically any data team.

Today at Bigeye we're focusing on data quality, which from among a number of things that we worked on at Uber was one of the most impactful areas. I know it's something that a lot of teams still struggle with and can have bottom-line impact for businesses. So that seemed like an important problem to go solve.

Why are you now raising a Series A for Bigeye?

Kirwan: We spent the first year or so of the company's life just sort of heads down in product-building mode and working with early design partners. The traction that we saw from the market when we finally started to commercialize was strong and the Series A allows us to fully embrace that.

From a logical standpoint, if society is moving to this model, where data is feeding things directly, there's going to be a need there to also automate the detection of problems with the data that feeds all of that stuff.
Kyle KirwanCEO and co-founder, Bigeye

Increasingly, modern businesses are using data directly in something that impacts a line of business. It's not just that there's an analyst looking at data and interpreting it and coming up with insights; data is actually being fed directly into things that users touch and feel every day. It's being used to route support tickets and it's being used to recommend products that you might want to purchase.

From a logical standpoint, if society is moving to this model, where data is feeding things directly, there's going to be a need there to also automate the detection of problems with the data that feeds all of that stuff.

So we went out to raise the Series A, because we were at this nice … zone of having had enough commercial traction that it was an appealing prospect to investors.

How do you measure and define data quality and freshness?

Kirwan: If you have access to data lineage, that is helpful information when diagnosing issues.

Freshness, I would say for sure matters since a lot of a lot of applications for data depend on knowing if they're being fed with recent information or not.

At Bigeye, we cover a range of different operational data quality issues, including formats, outliers, distribution, duplication and things like that. We collect signals about each of those different characteristics in the data sets that we monitor, and then that's how we report on data quality.

How much of the data quality process requires manual human intervention rather than being fully automated?

Kirwan: What we see a lot of our users do is they use the automation that we provide as sort of a first pass to get really broad coverage on all of their data sets. Then they can come in and layer in human expertise later in the areas where it matters the most.

The analogy I like to use is, 'If I gave you a broom, and put you in a dark room and told you to please clean the room up, how would you get anything done?' The first thing you'd want to do is turn on the lights.

So we like to think of it as like we're flipping on the light switch first and then the human with the broom can see what needs to be done.

Editor's note: This interview has edited for clarity and conciseness.

Next Steps

Syniti boosts DataOps chops with DMR merger

Superconductive raises $21M for open source data quality

Dig Deeper on Data governance

Business Analytics
Content Management