Imagine you... could simply have a conversation with your entire VC portfolio.

How AI can turn investments into an intelligent network - AI in venture capital

A VC with 20, 30 or 50 investments possesses an enormous amount of knowledge, yet has an astonishing problem in actually using that knowledge.

Every month, new figures, investor updates, board decks, financial plans, forecasts, meeting notes, and information from personal conversations are added. A large portion of this is evaluated, discussed, and filed away somewhere, while knowledge about individual companies remains scattered across different systems and, above all, across different people.

The real value may not lie solely in the information about each individual company, but in what could be discovered by considering the entire portfolio together and using AI in venture capital.

So what if an investment team could actually ask its portfolio:

• Which of our investments are expected to require new capital within the next nine months?

• Which companies are currently experiencing faster cost increases than revenue?

• Which of our companies are currently looking for similar talent?

• Who already has experience with a market into which another investment is currently planning to expand?

• Which companies work with the same external partners and could negotiate better terms together?

• Where is there knowledge within our portfolio that is currently needed elsewhere?

This is precisely where AI becomes truly interesting for venture capital, because then we are talking about much more than just faster reporting.

We're talking about portfolio intelligence.

A portfolio is actually a pretty powerful network.

VCs talk a lot about their network, and indeed, a good portfolio holds enormous potential.

There are founders and executives with very different experiences, specialists, contacts in specific industries and markets, knowledge of internationalization, recruiting, technology, production, sales, regulatory issues, and of course a large amount of experience about what has worked and what has not worked so well.

In reality, AI is surprisingly rarely systematically linked in venture capital.

A founder might spend weeks solving a problem that another company in the same portfolio already faced a year ago. Three companies might simultaneously be searching for a good enterprise sales lead in France, several might be negotiating independently with the same software vendors, and elsewhere a market is being developed where someone in the portfolio already has excellent contacts.

AI could make precisely these connections visible.

An intelligent system could, for example, recognize that Company A is currently seeking support for entering the French market, while Company C established a sales structure there two years ago. Or that six subsidiaries use the same service provider, resulting in a completely different negotiating position.

Suddenly, the value of the portfolio no longer consists solely of the individual holdings, but also of the connections between them.

Shared resources could arise from actual need.

Many VCs already offer their portfolio companies support in areas such as recruiting, legal, finance, marketing, PR or business development.

Things get interesting when you build these resources based on what is actually needed in the portfolio.

It might turn out that eight companies are currently experiencing difficulties with recruiting. In that case, a central talent resource could be more effective than eight separate solutions.

Perhaps several companies are expanding into the same market simultaneously and could share local contacts, experience, or even infrastructure.

Perhaps the same sales problem arises in five different places, and it is worthwhile to develop a common format from it.

Perhaps twelve companies are buying the same software and no one has yet thought about negotiating together.

AI could continuously recognize such patterns, thereby creating a completely different basis for how a VC organizes its portfolio services.

I find this particularly interesting because it makes Shared Resources look less like an additional offering from the fund and much more like they arise from the real needs of the companies.

At the same time, transparency within the fund is changing.

Internally, knowledge is often more fragmented than it initially appears.

Partners know certain companies particularly well, principals and associates have different information, finance considers different key figures than the operating team, and much of what is relevant for assessing an investment ultimately lies in conversations and thus in the minds of individuals.

A shared intelligence layer could generate significantly more institutional knowledge.

A partner could see at a glance which investments currently require special attention. Finance could analyze runway and future capital requirements across the portfolio. Operating teams could identify recurring issues and where joint support would be beneficial. New team members could much more quickly grasp how an investment has performed and what decisions have been made in the past.

This is an important point, especially for growing funds, because with each new investment not only does the portfolio grow, but also the amount of knowledge that needs to be available somewhere within the organization.

Reporting could also become significantly more intelligent for LPs.

The same database opens up another interesting possibility.

Limited partners (LPs) want to understand how a portfolio is performing, where opportunities arise, which risks become relevant, and how the fund manages its investments. To achieve this, information from various sources is regularly gathered, analyzed, and compiled into reports.

A well-designed portfolio intelligence system could generate different perspectives from the same data.

An investment partner needs a very deep operational view of individual companies. Finance is interested in different information. Operating teams, in turn, are interested in something else. A limited partner needs a condensed perspective on performance, development, and risk.

This would make it much easier to organize transparency, both internally and towards external stakeholders.

Of course, this includes a very clear data architecture.

The crucial question is not whether as many people as possible can see as much information as possible, but rather who needs which information for which decision.

Financial data, personnel matters, strategic decisions, customer information, and planned financing rounds all have different levels of sensitivity. Therefore, an intelligent system must be built from the outset with roles, permissions, and clearly defined information levels.

What information is the responsible investment team allowed to see? What information can be evaluated fund-wide? Which anonymized data is suitable for benchmarking? What can be shared within the portfolio? What information is relevant for limited partners?

These questions are at least as important as the choice of technology.

The real advantage goes far beyond efficiency.

Of course, AI saves time here. Reports can be generated faster, information can be found more easily, and trends can be identified earlier.

However, I believe the greater leverage lies elsewhere.

For the first time, a VC could truly leverage the knowledge, experience, and resources of its entire portfolio as a system.

Fifty investments thus become 50 companies with their founders, employees, contacts, markets, experiences, technologies and skills, which can be intelligently connected.

And the larger this network becomes, the more interesting it becomes.

This could eventually become a real argument for founders when choosing their investors.

Because in addition to the question of how much capital a fund invests and who they know there, another question could become relevant:

What network and knowledge will I actually gain access to if you invest?

Perhaps that's the more interesting AI question for VCs.

In recent years, VCs have talked a lot about which business models are emerging through AI and which companies they should invest in.

I am now at least as interested in what happens when a fund applies AI to itself.

If he better understands what is happening within his investments, recognizes connections earlier, makes knowledge accessible, organizes shared resources effectively, and can give different stakeholders exactly the transparency they actually need.

Then at some point we won't be talking about portfolio reporting anymore.

We are talking about a portfolio that functions as a network, retains knowledge, recognizes connections, and becomes smarter with each new investment.

And that's exactly where it gets interesting.