Data-Driven VCs: Who Is Using AI To Be A Better (And Smarter) Investor


I. A Data-driven What?

The ones of you who know me are very well aware that if there is something which has sort of obsessed me for the last few years, this is definitely how to use analytics and AI to improve the venture industry.

While I tended to focus on scouting and evaluation, I learned that AI can be also used to spot general trends, identify market gaps, improve VCs portfolio management, better match co-investors and deals, gather intelligence on competitors’ landscape, identifying potential acquirers, and improve pricing models.

I have been thinking about those issues for a while now (and stay tuned because I will post over the next few months my latest research in this field), and did already write on the importance of using AI in VCsummarized some academic research on this topic, and generally wrote about AI investors and accelerators.

I am clearly not the first one who is arguing for a data-driven approach to investing, and I thought it would have made sense to write a short post on other same-minded investors that I got to know or heard of.

Knowing exactly what they do is quite cumbersome without having inside information, so I am simply reporting public knowledge as well as what I read and heard on venture funds using AI to some extent (in alphabetic order):

  • 645 VenturesSeries A investor, they follow a data-intensive approach that mostly helps them in deal sourcing and evaluation (and they have a fairly specific metrics-driven process to investGrowth Seed in a bunch of different sectors). They also seem to automate many of the traditional VCs manual tasks;

  • Ardian: a world-leading private investment house, they are enhancing their AI capabilities through partnerships with startups that can collect and analyze unstructured market data;

  • Connetic Ventures: their data analytics platform Wendal collects, analyzes, and ranks startups, and support them in the due diligence process speeding it up to 8 minutes;

  • Correlation Ventures: probably the first real data-driven investor, it reaches a decision on whether to invest or not in 2 weeks, plus other 2 for extra due diligence. They only do co-investments in the US and do not take board seats;

  • EQT Ventures: with more than €500M AuM and equipped with an AI system called “the Motherbrain”, they have done more than 20 investments in less than two years. Apparently, most of its backbone is based on convolutional neural networks;

  • in addition to having a public dashboard to spot internet startups, they have been playing with analytics since 2012 to inform their investment process;

  • Fyrfly: there is not much information publicly available on their process, but this is another fund claiming a data-driven approach as part of their foundational principle;

  • Fly Ventures: the recently closed a first €41M data-driven fund to do small ticket size investments (up to €1M) and have invested in companies like Bloomsbury AI, recently acquired by Facebook;

  • Follow[the]seed: a post-seed global algorithmic VC, they have developed two data-driven methodologies (one B2B and one B2C) to simplify the investment process;

  • Georgian Partners: one of the most prominent Canadian VC, GP is not simply looking at AI for improving their investment process, but they also put machine learning at service of his portfolio companies (e.g., they developed a differentially private machine learning software for their ecosystem);

  • GV: everyone seems to know GV (formerly Google Ventures) is using AI and machine learning to inform their investment process, although almost no one knows exactly what they are doing and how;

  • Hatcher+: they use a data-driven approach to offer their partners quality analysis and opportunity scoring. They identify early-stage opportunities and created what they call a Resilient Investment Model;

  • Hone Capital: the Palo Alto-based US arm of CSC Group, it partnered with AngelList to create their proprietary model;

  • InReach Ventures: led by Roberto Bonanzinga and Ben Smith, InReach has quickly built a name as the software powered house able to scout early-stage European startups even before others VCs realized they need funding (see what happened with Oberlo, for example). Apparently, it took them two years and £5m+ in investments to build their proprietary software;

  • Nauta Capital: B2B software-focused European VC, they recently brought in a few very good software engineers and data scientists with an ambitious roadmap in mind. In fact, they are working on a prediction engine (that assesses the probability of a potential investment to be successful), a dealflow engine ( which automatically gathers and analyzes data of potential investment opportunities), and a dynamic reserves planner (that calculates the optimal distribution of reserves for follow-on);

  • NorthEdge Capital: private equity investor that developed a platform that can identify new investment and buy-and-build opportunities in the area by analyzing regional-specific companies;

  • Origin Ventures: they timidly admitted they built their own scoring software to assess the quality of startups, but there is not much out there to back this claim;

  • Redstone: renowned VC that introduced the well-known “VC-as-a-service” model, it did hire a few years ago an incredibly talented scientist (Stefano Gurciullo) who is, using Redstone’s words, “building technologies that help them invest based on evidence and on a quantitative understanding of innovation”;

  • Right Side Capital Management: with more than 800+ pre-seed investments done so far, they make small investments ($100k-$500k at valuations of less than $3M);

  • Scale Venture Partners: their Scale Studio is a platform that allows startups to compare their progress to similar startups across a handful of key business metrics;

  • SignalFire: the firm run by Chris Farmer does not only use analytics to pick the right companies but also to help them grow by providing market intelligence and talent matching services;

  • Social Capital: led by Chamath Palihapitiya, the firm is better known to have started the Capital as a Service (CaaS) concept, and more recently they created an analytics due-diligence tool (which is hosted on their webpage) to help them invest in early-stage companies and identify trends in customer cohorts;

  • Switch Ventures: they are using a mathematical and predictive approach to sourcing, but there is not much around to better explain what they are in fact working on;

  • Ulu Ventures: they use Decision Analysis to inform their investment decisions. More in detail, they use it to create market maps, assess risks, quantify uncertainty, perform sensitivity analysis and eventually compute the risk-return profile of a potential investment;

  • Venture/science: a quant-driven VC led by Matt Oguz, it uses AI and decision theory to compute the risk associated with different attributes such as team completeness, vision, etc.;

  • WR Hambrecht VenturesThomas Thurston is the key man behind WRH (and Growth Science, its sister tech company) who is advocating for the use of data science to guide growth investments for years now.

From the original list of about 10–15 funds made a year ago, we now have more than 25 funds that are using AI in different ways. Even though this increase may look like a drop in the ocean of the venture industry, it seems to me something extremely relevant that may change, in fact, the way we think about investing.

II. Building or Buying It?


Is this the whole story? Not quite.

Even though I started this post (and my search) focusing on venture funds that use AI in different ways, I eventually discovered that VCs are not the only players in this niche industry. In fact, there exist several startups and tools that I think are worth mentioning for the sake of completeness because they are trying to democratize VC investors’ skills:

  • Aingel.aithey have recently filed a patent for a machine learning system that scores startups and founders and also matches the companies to the most suited investor;

  • Capital Pilot: another service that facilitates the match between companies and investors;

  • Crunchdex: a new company that is focusing its attention on identifying the fastest growing startups;

  • Kognetics: they have a proprietary framework to identify interesting deals and offer extra insights on trends, markets, and competition;

  • Preseriesthis is another fully automated solution to discover and evaluate startups, which it also has a voice interface (through Alexa);

  • Radicle: their proprietary software can be used to detect novel interesting sectors, and I believe they have something to say also on new ways to evaluate startups;

  • Rocket DAO: a decentralized crowdfunding and startup evaluation platform, they help companies and investors to better match (still in beta);

  • Valsys: they provide professionals with the tools they need to make data-driven decisions in valuation and estimation processes. They focus however more on a later stage.

This is likely only a partial list, but it conveys and bolsters the point mentioned above: having an AI-driven investment engine is becoming a trend, and we should expect more of those solutions in the future.

It is also interesting to note that there are more funds pouring money into these engines development than companies selling those systems as a service. In other words, VCs seem to prefer building to buying when it comes to intelligent software for their own internal use. Intuitively, this is paramount to create a moat and a competitive advantage with respect to other investors, but it is also true that this could segment the market and polarize it: in fact, while bigger funds may have the resources to invest into building their own platforms, this may not be true for smaller funds, and this could also result in wrong signaling to LPs and potential deals (e.g., if you buy a software rather than building, you may be considered to be a second-class investor).

Furthermore, a last comment. We listed so far funds and mainly software companies that are offering different types of AI services. These are not the only two options though. There are intermediate alternatives such as the one provided by Clearbanc and 20-Min Term Sheet, where they use algorithms to review the startup’s marketing and revenue data and decide whether to grant a loan in about 20 minutes. Similar capital-as-a-service offers are provided by other companies such as BlueVineLighter CapitalCorl, always with an automated process that speeds up the investment decisions.

III. Conclusions


I am pretty optimistic about data-driven VCs being the future; although I had a finger in the pie for long enough to understand that it is not trivial as people believe (or tell). This is why I am spending a lot of time thinking and working on how to push it further, and I will release very soon a couple of new interesting posts on practical research I have conducted on the topic.

I do not believe the future of the industry is to be fully automated and VC is and always be a people business. On the other side though, it sounds astonishing that algorithmic thinking has not permeated so far the way investors work on a daily basis.

Finally, I spent time researching and talking to many of those people, but it is also very likely that I might have misunderstood something or missed someone out there who are working on similar approaches. If so, please feel free to reach out!

Francesco Corea, Ph.D., contributor, is a complexity scientist and tech investor who is mainly focusing on science-driven companies in high-social-impact verticals including life sciences, energy, and artificial general intelligence. Francesco has a background that ranges from economics and finance to applied machine learning, and he's been working on a variety of different data problems over the past few years (e.g., sentiment analysis, fraud detection, behavioral science, etc.). Currently, he is working with a few AI companies as well as an emerging VC fund. Dr. Corea holds a Ph.D. in Economics.