18 months in, CME Ventures has sharpened its focus
Since CME Group launched its startup investment arm CME Ventures 18 months ago, it has refined its investment focus and taken stakes in six companies. It flushed out one investment target from an unlikely source: an internal CME human resources program. Rumi Morales ⇒, executive director of CME Ventures, explains.
Q. How has CME Ventures evolved?
A. It’s evolved insomuch that we’ve been able to define with more precision the types of companies we’re seeking to invest in. We’re looking at five general categories: security, real-time payments, what I call precision big data, machine learning or deep learning, as well as next-generation computation technologies.
Before, we were looking broadly at technologies that affected products, technologies, users or platforms. Our overall mission has not changed, which is to be able to invest as effectively as we have. The aim is to look three to five years out toward transformational innovation.
Q. Nervana Systems was your most recent investment. What about them got your attention?
A. This is a special one for us. We’d been looking in the deep-learning and machine-learning space for a while. It was a bit hard to find the right target at the right time. We have an internal human resources program here in the CME in what we call a talent exchange program. It allows employees from around the companies to participate in special projects.
The ventures team hosted one such program whereby we invited 10 people across the company to go through our early-stage pipeline and do their own research to vote on their favorites. Nervana was one company we threw into the pipeline. For this class they all voted for Nervana. So we looked into it. We do this talent exchange every quarter. So the next quarter we threw in Nervana again, and again, it won in the same program.
Q. What ultimately got you to invest in Nervana?
A. It fit our investment thesis and criteria. We had some assistance from the University of Chicago to analyze their technology, and we met with the management team and visited them in San Diego. They’d already done remarkable work in image recognition and natural language processing.
What is great is they are making this scalable. They are creating what we would say is the full stack. It’s the way for people to develop their own deep learning applications based on Nervana’s software and hardware and algorithms. We see it as a foundation technology.
Q. You recently said that you want to see startups communicate better with companies. How did you mean?
A. You can imagine as a corporate VC, we see a number of startups come in and say, “I’ve got a company that can solve problem X.” And I say, “That’s really not a problem for us.” They will say, “Well, we can solve for Y,” and I say, “We’ve already solved for Y.”
I don’t blame startups for not knowing this. Many times corporations do not share the innovations that are going on in their own buildings. But corporations innovate too, a lot. It would behoove us to be able to talk about that some more.
For everyone that says, “Silicon Valley is coming,” or “Watch out, big institutions,” what I like to say is that through the Board of Trade we’ve been here since 1848. I ask startups, “Is your company going to be around in 170 years, and if so, what will it look like?” Please don’t discount large corporate innovation. The work that the Ventures group is doing is another iteration of that.
Q&As are edited for length and clarity.
Kate MacArthur is a freelance writer.
Copyright © 2015, Chicago Tribune
Source: 18 months in, CME Ventures has sharpened its focus
Via: Google Alert for Deep Learning