DataSift made a name for itself as a company that took large unstructured datasets — such as anonymised firehoses from social media services like Facebook (and previously, Twitter) and ordered them in a way for enterprises and brands to get a better idea of consumer preferences and other insights. Now Nick Halstead, the person who founded and led DataSift but left the role in October 2015, has started a new company that wants to provide big data insight directly to the enterprises themselves — based on their own data.
CognitiveLogic, as the new, three week-old startup is called, is today announcing a seed round of $3 million to kickstart the business. The funding comes from influential U.S. VCs Upfront Ventures and IA Ventures.
While there are a number of analytics startups out there helping leverage companies’ own internal databases in better ways, CongitiveLogic is offering a different twist on the idea: it’s concept is to join up data from different enterprises to analyse it. “We want to join data between multiple companies without either company having to send the raw data, but gain insights from combining those datasets,” Halstead says.
There is a privacy element here that CognitiveLogic is trying to address: “Today many companies who sell data, and tell the public that they are PII compliant, really are doing the bare minimum,” Halstead says. The result? Data breaches.
“My work at DataSift has shown me many new technologies that will allow us to build solutions that allow companies to do research but without exposing individuals private information,” Halstead says.
The first verticals that Halstead says the company will target are government, health and finance.
Apart from the attention to privacy, I’d argue there is another notable aspect to Halstead choosing to work on solving an enterprise-based data challenge: DataSift’s business essentially rests on large, third-party platforms playing ball — that is, providing DataSift with access to their data firehoses so that DataSift could work its data science magic on them. The problem with that is these companies might turn around at any point and decide to try to do that analytical crunching themselves — which is essentially what Twitter did when it cut off DataSift from its own firehose.
That leaves companies like DataSift in a precarious position longer term — an enterprise-grade version of the same conundrum that impacts third-party developers that rely on other large platforms. CognitiveLogic is built without needing to worry about contingencies related to this: the main issue for it will be simply to provide interesting enough data analytics services to convince enterprises to work with it directly — much like many other SaaS businesses.
CognitiveLogic today has four people but plans to ramp that up to 10 in the next six months.
It’s interesting that an entrepreneur who has spent the last eight years building and working for UK startups has turned to the U.S. for funding.
Halstead says that this is because — despite the many reports of funding being more difficult to raise today than, say, years past — “the U.S. is still the best place in the world to raise funds with investors who both understand the data market, and also those who can follow on when required.”
“I’ve worked with Nick on solving complex data problems for more than six years. He is a true visionary in the fields of ETL, data normalization, real-time data streams and data privacy, and I’m excited to be his partner once again,” said Roger Ehrenberg, Managing Partner of IA Ventures, in a statement