Menu Close

Two Sigma acquires Hivemind, strengthens data-driven investing

Were you unable to attend Transform 2022? Check out all of the summit sessions in our on-demand library now! Watch here.


Ed. note: Corrections made on 8/24/22 to remove references to “Two Sigma Investments,” “hedge fund,” “fund,” and revise sentence saying “Crux, acquired in 2018…” to instead say, “Crux, a strategic minority investment….”

Two Sigma, a financial services company focusing on data-driven quantitative investing, has announced the acquisition of Hivemind Technologies’ software platform to strengthen its data-handling capabilities. The terms of the deal were not disclosed.

Founded in 2001, New York-based Two Sigma combines rigorous inquiry and data analysis with a variety of technological methods, including artificial intelligence (AI), machine learning (ML) and distributed computing, to inform its trading strategies. The company claims its data-driven approach helps solve the toughest challenges in the investment management and venture capital space. It gets over 20 petabytes of data from more than 10,000 sources every year and currently has more than $60 billion in assets under management.

Meanwhile, Hivemind Technologies, headquartered in London, is an IT consultancy that combines automation of its platform and distributed human intelligence to create valuable, bespoke datasets from messy or unstructured sources. The platform aims to bring control and efficiency to data processing, making it more auditable and accurate.

Event

MetaBeat 2022

MetaBeat will bring together thought leaders to give guidance on how metaverse technology will transform the way all industries communicate and do business on October 4 in San Francisco, CA.

Register Here

How does Hivemind software fit in?

With this deal, Two Sigma plans to enhance the data preparation, data workflow, and machine learning capabilities of its data engineering platform. The company said it will integrate Hivemind’s offering to glean the ability to distill semistructured and unstructured data sources into high-quality, machine-readable datasets.

“Hivemind provides us with a framework for inserting the various forms of labeling into unstructured data. It brings more automation and accuracy to the preparation and onboarding of data, to enable cleaner data and better research. This, in effect, frees up our data science and research teams to focus on research rather than data prep, and our engineering teams from supporting non-standard datasets and rote data cleaning,” Carter Page, head of data engineering at Two Sigma, told VentureBeat.

“In financial services, we have to make sense of the world and markets through data that is often complex and not machine-interpretable. Hivemind helps us to filter more complex data to make it useful,” Page added.

Two Sigma’s end goal: Better investing

Ultimately, the integration will amplify Two Sigma’s ability to ingest data quickly and utilize it more effectively for investment decision-making. In fact, the company has been at it for years, working to embed greater efficiency and accuracy into various aspects of its data pipeline, from cleaning to mapping to creating unique datasets.

“Hivemind is one of the pieces of our larger data pipeline ecosystem, with all of our tools working together as part of a single platform to move data efficiently through the computation and analysis that drives decision-making in our organization,” Jeff Wecker, CTO of Two Sigma, told VentureBeat.

Other pieces in this ecosystem are Crux and Liberty Source, respectively. Crux, a strategic minority investment, augments Two Sigma’s ETL activities and helps improve data quality and governance in structured data while providing access to a growing catalog of vendor datasets. Meanwhile, Liberty, acquired in 2021, provides data workflow management services, including entity mapping, data validation and curation.

“Our goal is to obtain the data we need, evaluate its value, and use the benefits of that research automatically in all of our investing and trading businesses, including more recently private investing,” Wecker said.

“While we’ve been developing these capabilities over the last 20 years, we still have much to do to get it right. We have recognized that the part of the data world that is most underserved is end-to-end pipelining. We want to eliminate all the friction from research to production.”

To achieve this mission, Wecker emphasized, the company will continue the search for external solutions that provide the tools needed or build them internally.

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn more about membership.