Shaping the Future of Venture Capital with Data and AI

Max Fleitmann
Founder of VC Stack
Shaping the Future of Venture Capital with Data and AI

In the world of data-driven venture capital, technology and analytics create a new era of investment opportunities. A data-driven fund can harness the power of data and machine learning to get in early on companies with the potential to disrupt their industries, streamline internal operations, and support portfolio companies. In this article, Pedram Birounvand, Head of Data Management at EQT, and Maximilian Fleitmann, founder of VC Stack, take a deep dive into the workings of a data-driven fund - from how data is harnessed to what roles comprise a data team.

Background

Launched in 2016, EQT Ventures is the venture capital arm of EQT Group, a global private equity firm with just under 1,000 employees. EQT has had huge success following the trend of digitization and big data. EQT Ventures recently closed Fund III with 1.1bn EUR in total commitments - Europe’s largest VC fund committed to early-stage tech start-ups. As the Head of Data Management, Pedram ensures that EQT Group has good, solid data platforms for internal operations with the right data structures to ensure processes are as efficient as possible. 

How is data used at EQT?

EQT built its own AI-driven system to help identify opportunities that would otherwise go unnoticed or get picked up too late. Their AI, referred to as Motherbrain, enables deal sourcing and helps make investment decisions. The algorithms use convolutional neural networks to map connections between data such as app store downloads, information on previous investors, website traffic, founders’ resumes, and more. It has allowed the firm to reach out to promising start-ups well before they are inundated with offers from competing firms and with a high degree of confidence. Their website notes that nine investments were fully sourced through Motherbrain - investments that would otherwise not have been identified - with the earliest ones proving to be successful so far. 

Likewise, it has also helped with internal operations by helping partners prioritize start-ups to invest their time in and preparing for meetings, such as by providing summaries of past interactions. To get to this stage, Pedram noted that EQT 

“invested heavily in data scientists to create connections in unstructured data to find good companies… [this] required good data science and engineering knowledge… [and] invested in a state of the art data strategy [including a] cloud data platform - GCP [Google Cloud Platform]”

Pedram added that data is also used for value creation at portfolio companies. This includes:

“helping portfolio companies with market data, identifying the macro situation in a region, sharing … [and] data with other portfolio companies that are not in direct competition with each other”

Portfolio companies are further supported with self-service data products. For example, in niche segments like ESG, EQT is building a product that a portfolio company can access to compare its ESG performance against benchmarks. This is in addition to additional ad-hoc data requests that EQT can support on a case-by-case basis.

How should data be used in small vs. large funds? 

Pedram noted that while data is critical for all funds, the level of sophistication in how it is used varies: 

Small companies should use off-the-shelf products … to be successful in building your custom product, you need to have 4-5 people working on it - this is a big investment for a small fund”. 

Pedram added that, on the other hand, if we are looking at a bigger fund, say with $1bn+ in AUM, it is worthwhile investing in an in-house team. Using a proprietary tool will give a competitive advantage over those using an off-the-shelf tool. 

Modern AI and natural language technology will give these funds an edge. Pedram added that ChatGPT has revolutionized how we interact and communicate on the internet. Funds can now use ChatGPT to get more structured data out of it. 

On a side note, at VC Stack, we took a deep dive into how ChatGPT can be used for investing. You can find some investing use cases and inspiration here.

What advice do you have for funds looking to adopt new data tools?

An easy way to assess suitability is to use the tools on a trial basis. Pedram commented that this is very much possible with today’s cloud software. He added that when negotiating with vendors, “get a free trial or something for cheap so that you can test it and opt out of it if you are not happy with it”. 

It also helps to have an expert question the vendor and cut through the jargon by asking the right questions and judging the responses.

If I am a GP, what tips do you have for hiring a data team?

Pedram noted that technology to manage data had evolved significantly over the last 20 years, and as a result: 

“The roles to hire for are changing all the time… [further] the same role would have significantly different responsibilities based on the company they would work for. For example, the responsibilities of a data engineer at Spotify are very different to one at EQT.” 

Accordingly, it is essential to know the exact responsibilities of each role in the team is essential. That said, Pedram’s guidance is to hire for the following roles:

  • Data engineer: while they won’t build custom code, they can write SQL code in a simple way so that a business person can understand it.
  • Data scientist: only hire for this role if they are qualitative datasets to work with.
  • Analyst: their role is to understand the data, make sense of it and expose it to the broader business.

How do you envision the future of investing?

Data is critical to investing. Pedram commented:

"Venture funds need to find companies in much earlier stages and invest in more deals than we do today… [having] computers and machines do the leg work [means we can] do a lot [more] deals [leading to more] investment in earlier stages and more niche products.

Concluding Thoughts

As a data-driven business, EQT has built its proprietary system, Motherbrain, to identify and invest in at least nine start-ups. In this interview, Pedram shared his thoughts on the role of off-the-shelf products for smaller venture funds and the importance of trying them out before incurring significant business expenses. We also looked into how hiring for a data team is not just about filling particular roles but identifying the responsibilities that need to be fulfilled. As we look to the future, data and AI capabilities will be vital in identifying the most promising venture capital deals - it would be a significant loss to miss this industry shift.