Cognitive Business: AV Shares Their Perspective on the AI World

In this Cognitive Business interview, we speak with Navid Alipour, Managing Partner at Analytics Ventures (AV), a venture capital firm focused on internet of things, machine learning, and artificial intelligence startups, such as CureMatch and CureMetrix.

Two things set them apart: their business model of incubating their own ventures and their owned AI lab with world-class talent and proprietary algorithms for time sequence data sets and image analysis.

Can you tell me a bit more about the Analytics Ventures fund?

We help build breakthrough ventures that, in some cases, would not exist without us. We partner with domain experts, scientists, academics, and entrepreneurs to help take companies from inception to Series A. There are fantastic VCs who can help scale a company from 20-30 employees to 1,000+. We are focused on getting them through that valley of death that so many companies fail in.

Analytics Ventures has been around for a little over four years. Until last year, we did not have investors, so we put our own money to work. As we were approached by scientists with a certain domain expertise, we started co-founding companies deploying the venture studio model like IdeaLabs in Pasadena or ScienceLabs in LA. In this process, we had co-investors ask how they can get access to founding shares, and therefore, we decided to raise a fund.

Now, Analytics Ventures Fund I LPs get through the fund direct ownership of founding shares as the Fund is a co-founder alongside other founders who could be individuals or other corporate partners. I like to say we are a Venture Fund, operating under a venture studio model, focused on any vertical deploying AI, machine learning, or deep learning to glean value from data. I am fortunate to have amazing partners in Blaise Barrelet and Andreas Roell; one a native of France and the other Germany who are amazing operators and entrepreneurs.

Blaise founded WebSideStory which was one of the first and most successful Internet era companies founded during the late 90s. They were the "pulse of the internet." The company IPOed and ultimately got acquired by Adobe for $1.8 billion.  Andreas has founded several companies and sold to companies such as ESPN to Rotana out of the Middle East, where he built the Arabic version of Hulu. I like to say he brings the German operational efficiency to our process.

How does the firm make decisions on AI investments?

For starters, all three Managing Partners have to be unanimous in our decision to incubate a new venture.  Some considerations include:

• Do we have access to data sets?
• Do we have a partner that has expertise in a domain to allow us build a breakthrough hypothesis?
• Does it align with the algorithm and scientist capabilities that make us unique?
• Is the path to market validation within a 6-month roadmap?
• Can we establish unique IP?
• Is the TAM (total addressable market) large enough?

Is there any area where you think your firm adds unique or disproportionate value to AI startups?

We have an AV Lab which is an AI lab with some brilliant talent. Companies don't have to re-invent the wheel and hire their data scientists out of the gate. We can bring the talent, not just an investment. As a result, our ventures have a significantly shorter and lower capital requirement around the areas of implementation of AI. Also, we have developed a complete operational framework based on our experience of having created over 14 successful ventures over the years. This framework strips away all of the noise early stage ventures face and allows them to execute in a very clear and succinct fashion. It is applied to all of our ventures, where they also benefit from preferred access to service providers that we have established.

What are your return expectations for an AI startup?

That is a difficult question to address. With our venture formation model: we start companies and own founder shares through the fund. We can have a very nice ROI still even by hitting singles and doubles to use a baseball analogy. That said, we have to believe like with any investment, that the pain point must be significant, the total addressable market must be large, and the team must be the right team to get the job done.

Do you have an ideal AI startup exit strategy?

As early stage investors, an exit can certainly come from one of the Fortune 500 companies, or it could be through private companies. We can also take money off the table as many founders and early stage investors do on subsequent later rounds that can provide us a nice return.

What is your advice to Fortune 100 tech companies looking to acquire AI startups?

Contact us. Also, get involved with the investors in the space, the institutions, and put up scholarships and programs at the high school level even! The fear in academia right now is the big companies are attracting all the talent, and there is a shortage of professors to teach the next generation. The Fortune 100 in all verticals should foster AI education at the high school level on up. To paraphrase Mark Cuban, “if you are not learning about implementing AI in your respective vertical, you will be a dinosaur in 2-3 years.”

Also, a core problem with the high popularity of AI currently is the fact that most companies all the way to Fortune 100 have limited depth of understanding the technological differentiators of Artificial Intelligence. Just by labeling something artificial intelligence does not mean that its technology is truly based on AI. Also, our belief is that many of the current AI technologies are built on very shallow first mover environments and the limited experts available outside of the Googles and Facebooks of the World. Thus, we advise all companies thinking about an acquisition with an aspect of AI to fully understand the sustainability of the AI technology used. We predict that many of today's AI leaders in verticals will be swallowed up by much more sophisticated solutions that have the ability to work off sparse data sets, unstructured or noisy data while deploying transparent algorithms to allow for human guided oversight. All of these factors are currently available on a very limited basis.

On the flip side, what advice would you give AI startups looking for VC funding?

The advice I gave to the previous question applies on the flipside also to entrepreneurs. Have a full understanding how strong your underlying technology is in today's and tomorrow's marketplace. Just using, for example, an off the shelf IBM's Watson AI algorithm on top of your approach does not mean necessarily you are going to be great. An objective perspective will help you identify if your opportunity is short-term or long-term. If short-term, you might simply want to ride the wave of AI enthusiasm that exists in today's financial and corporate environment. There is nothing wrong to build something and sell off quickly. On the other hand, you can also have a short-term orientation to get to market as an early mover and use the advantage to build out your longer-term, sustainable platform.

In either case, think clearly about your approach to AI. Build your ecosystem of experts in the field as the financial community and investors are getting up to speed very quickly at this point.

If you have something unique, reach out to investors in the field, especially those near you. Use social media to do your research as well on those in the space. If you are at an academic institution, check in with their Tech transfer office.

How can we get in touch with you?

Email us at info@analytics-ventures.com.

“Cognitive Business” is an interview series featuring awesome people in the Artificial Intelligence (AI) world. Written by Lolita Taub and written for business people.