Sifflet is raising money to expand its data observation platform

by Ana Lopez

Organizations dealing with large amounts of data often struggle to ensure that data remains of high quality. According to a questionnaire of Great Expectations, which creates open source data testing tools, 77% of companies have data quality issues and 91% believe it affects their performance.

In light of that, it’s not surprising that things are quite healthy for vendors selling data observation services and software that help an organization understand the health and condition of their data. Last year, in the space of a week, three data observation companies alone — Cribl, Monte Carlo, and Coralogix — raised more than $400 million.

Suggesting that the market is not yet oversaturated, another data observation startup secured venture capital this week: Sifflet. Today, the company announced that it has raised €12 million (~$12.7 million) in a Series A financing round led by EQT Ventures with participation from existing investors.

Sifflet was founded in June 2021 by Salma Bakouk, an ex-VP of Goldman Sachs in sales and trading. She teamed up with software engineers Wissem Fathallah (formerly at Uber and Amazon) and Wajdi Fathallah to launch an MVP, which grew into a full-fledged data observation product.

“Sifflet is a data observation platform that helps companies build trust in their data,” Bakouk told in an email interview. “The platform sits above the data stack and provides a 360-degree view of the data assets.”

With Sifflet, companies can collect information across different layers of their data stack, from the data ingestion stages to transformation and consumption. The platform automatically checks data, metadata and data pipelines for indications that something is wrong, such as a sudden drop in quality.

Sifflet maintains a line to make it easier for data engineers to perform root cause analysis. As Bakouk explains, AI is at the center of this process.

“AI is being used in our monitoring engines, data classification and context enrichment,” she said. “Our models are pre-trained against various types of datasets from different industries and dynamics and are regularly re-trained when deployed to account for the specifics of the client’s environment and reduce any training bias.”

So, given the competition in data perception, can Sifflet reasonably compete? The investors clearly believe it can be done. A more objective measure is the size of Sifflet’s customer base, but Bakouk does not want to disclose this. She did, however, report that Sifflet counts brands such as Carrefour, Nextbite and ShopBack among its current clients.

“Sifflet’s approach is purpose-built to be inclusive for most data practitioners, both technical and non-technical,” Bakouk said. “In the current economic climate, where companies are faced with tough decisions, data-driven decision-making is the norm and data incidents are simply not tolerated.”

It’s hard to argue with that last point. Poor data quality according to Gartner costs organizations an average of $12.9 million per year. In addition, data engineers spend two days a week putting out bad data, a poll found from Monte Carlo.

“The economic slowdown is actually a great catalyst for data adoption. Businesses need to take uncertainty out of the equation when making tough decisions and data reliability is key,” said Bakouk. “In terms of the company’s position, we value capital efficiency and look for strategic ways to grow. Having a laser-sharp product vision from day one enabled us to be focused and fast in execution, avoiding costly pivots.”

Paris-based Sifflet, which has raised €15 million (~$15.85 million) so far, plans to step up its go-to-market efforts in Europe, the Middle East, Asia and the US and continue to invest in products and engineering. It currently has 28 employees and aims to more than double that number by the end of the year.

Related Posts