Gretel AI raises $50M for a platform that lets engineers build and use synthetic data sets to ensure the privacy of their actual data

DMCA / Correction Notice
- Advertisement -

Conversations about big data, machine learning and artificial intelligence are growing rapidly, along with conversations about privacy and data security. Now, a startup that is building tools to make it easier for engineers to implement both together is announcing a round of development funding to continue expanding its operations.

- Advertisement -

Gretel AI, which lets engineers create anonymous, synthetic data sets based on their real data sets to use in their analytics and has closed $50 million in funding to train machine learning models, a Series B that the company will use. to be taken to the next stage. Why development? The product — which is built as a SaaS product but can also be accessed via an API — is still in beta, but is targeted to be open for general availability later this year.

Anthos Capital is leading the round and Section 32 is also participating along with Greylock and Moonshots Capital. Greylock led the company’s previous round in 2020, and the startup has raised $65.5 million to date.


From what we understand, this latest round puts the company’s valuation at between $320-$350 million.

The idea behind using synthetic data sets is that it lets an organization remove the risk of leaking data that may contain personal information or other types of sensitive data. There are other solutions to address the same issue involving data encryption, although this can be a costly, time-consuming and resource-intensive approach that poses scaling challenges.

- Advertisement -

The sprouting for came from the three co-founders’ direct experiences, which they did over the years as cybersecurity experts at multiple organizations including IBM, AWS, NetScout and the US military.

“We’ve always found that using the right permissions with data was always a bottleneck,” said CEO Ali Golshan, who co-founded the company with Alex Watson (CPO) and John Myers (CTO). They could see that the long-term issue would be a growing need and priority for data privacy. “As the world moves from the web to a wider world of censors and IoT, we are transitioning into a world where people will unknowingly or unknowingly share their data. But humans are not for mining.”

As data engineers, their priority is to be able to work with data easily and quickly, but as citizens of the world, they were unhappy with the data security implications.

“Removing the compute bottleneck is the problem we’ve solved, and we’ve created high-velocity development,” he said. “But now we are running into a data bottleneck. AI is on a collision course with privacy. In this collision course, we must build the tools to fix it”.

Gretel’s opportunity is one that many companies targeting the enterprise market have taken to the world of digital transformation: many organizations now have large engineering operations working on applications to drive their businesses, but they still have access to the world. The biggest technology companies don’t have the firepower. So Gretel set out to create a toolkit that would allow any company to create anonymized data sets for themselves, similar to what big tech companies use in their data work.

The benefits of anonymous data go beyond simply replacing a synthetic data set for a real one; They can also be used to augment a data set, or to fill in gaps where real-world data may be lacking. Both of these are important components, especially in cases where data is needed to train the system, such as in the case of autonomous services, where you may never have enough learning data.

Watson previously worked at AWS (fun fact: when Amazon acquired its previous startup,, and he says that to this day Has secured early customers in areas like Life Sciences, Financial Services and Gaming. For more basic use cases, it can take at least 10 minutes to create a synthetic data set. In more complex applications – for example in genomic databases, this can take several days.

Relatively speaking, this represents “much less friction” for engineers, Watson said, compared to both other methods such as data encryption using techniques such as homomorphic encryption, or actually contacting third parties and storing datasets. Analog approach of obtaining permission to use . The latter can take six months or more, in cases where time is of the essence.

Emily White, President of Anthos Capital, said, “This significant Series B investment is a direct reflection of Gretel’s ambitious vision, rapid growth and positioning strength in the AI ​​industry as the standard bearer of devices that enable privacy by design. ” a statement. “Gretel’s ease of use, the scalability of its services, and the superior accuracy and quality of its synthetic data are much-needed solutions to simplify companies facing extremely complex legal and technical hurdles due to data privacy concerns.”

“Gretel gives data teams working in any framework or language the tools they need to create privacy by design in their existing workflows and data pipelines, greatly simplifying the process,” said Greylock partner Sridhar Ramaswamy. “I hear again and again from software engineers and data scientists about the value of Gretel’s offering. Its developer-first, The tech-agnostic approach to solving privacy issues is incredibly valuable to every business sector.”

- Advertisement -

Stay on top - Get the daily news in your inbox

Recent Articles

Related Stories