To Investors,
The other day I came across a post from Steven Bartlett about how there isn’t enough of a discussion about AI agents and how they are disrupting every industry. I wanted to discuss how these agents apply to the fields that I’m involved in, and even though I’ve tracked a lot of this activity for years, I’m continuously mind-blown at the information that I come across.
What are AI agents?
AI agents are software systems that perform tasks or make decisions by processing information and acting toward specific goals. Advanced AI agents can reason, handle complex tasks, and sometimes collaborate with humans. The best ones improve over time by learning from new data, algorithms, or experiences — much like how you and I refine our decisions with more knowledge and intuition.
According to a March 2025 report from Datagrid, adoption of AI agents among fortune 500 companies has grown 450% since 2022, and these agents are improving diagnostics in healthcare; assisting with inventory management in retail; they’re reducing costs and improving operational efficiency in manufacturing; they’re being used for precision farming; and obviously we’ve all seen the self-driving capabilities that Tesla and Waymo are unlocking. There are way more applications that we can discuss, but today I want to talk about how AI agents and AI more broadly is disrupting the private capital markets (i.e. venture capital and private equity).
I’ve approached this thought two ways:
AI for Internal Operations in a VC/PE firm.
AI as part of the investment decision-making criteria.
Internal Operations
Traditional venture capital and private equity involves investing in companies with high growth potential, typically in exchange for equity, with VC/PE investment firms playing an active role in guiding these companies while in the early stages. In the later stages of a company lifecycle, capital allocators may play a role in helping the companies transform their operations to improve their capacity and scale. An AI system replacing this model would need to replicate or enhance four key functions:
1. Deal Sourcing:
Purpose: Identify promising startups.
How It Works: The AI scans vast datasets to find opportunities, analysing industries, team backgrounds, early traction, and market trends.
Example Features: Real-time alerts for emerging startups or ranking systems based on growth potential.
2. Due Diligence:
Purpose: Evaluate the business’s potential and risks.
How It Works: The AI processes business plans, financials, and market data to assess viability and scalability.
Example Features: Automated risk scores, financial health projections, team competency analysis.
3. Investment Decision:
Purpose: Decide which businesses to fund.
How It Works: The AI uses predictive models to score startups against your firm’s investment criteria, recommending the best fits.
Example Features: Success probability forecasts or portfolio fit analysis.
4. Portfolio Management:
Purpose: Monitor and support invested companies.
How It Works: The AI tracks performance metrics and provides insights or recommendations for growth or exits.
Example Features: KPI dashboards, exit timing predictions.
An internal operations AI model would be data-driven, scalable, and capable of reducing human bias while speeding up processes traditionally reliant on manual efforts.
Let’s get practical. You’re probably wondering if there are any VC/PE investment firms that are actually using AI agents in their internal operations. The answer is yes, and many have been doing so for years…
EQT Ventures (Motherbrain)
EQT Ventures developed Motherbrain in 2016, an AI platform that proactively sources investment opportunities by analysing vast datasets, including startup performance, market trends, and founder profiles. It’s used for deal sourcing and due diligence, in partnership with their investment team’s decisions.
Correlation Ventures
According to an article from MLQ from 2021, Correlation Ventures uses a machine learning model trained on data from over 100 000 VC rounds to guide investment decisions. The AI analyses pitch decks, team experience, and board composition to predict success.
Hone Capital
Hone Capital is the U.S. arm of a Chinese VC. According to an interview with the managing partner, Hone Capital partnered with AngelList to create a machine learning model based on 30 000 deals. At the time of the interview, in 2017, it was claimed that the model evaluated 400 characteristics (e.g., founder background, funding raised, etc) to identify startups likely to reach Series A. In that interview it was reported that 40% of their AI-recommended companies raised follow-on funding, 2.5 times the industry average.
SignalFire
SignalFire uses a proprietary AI platform called Beacon AI, that, according to the firm’s website, tracks 2 million data sources to find investment opportunities, helps portfolio companies with a go-to-market strategy, and also helps them to recruit talent.
Bain & Company
According to a 2023 announcement on Bain & Company’s site, the firm developed a proprietary advisory model called Sage, which was built to help all advisory employees in the company with generating insights and business ideas.
Two Meter Capital
Founded in 2024, Two Meter uses AI to provide portfolio management services to family offices, funds, and corporate venture groups. Services include portfolio tracking/reporting, support for follow-on financing, and assistance with exit processes.
D3VC (AI Fund)
D3VC’s AI Fund uses a proprietary AI and machine learning model, trained on proprietary data, to identify patterns that indicate whether a company is on track for follow-on financing or other markers of success.
Those are 7 stimulating examples of how VC/PE firms are using AI as part of their internal operations to assist with deal sourcing, due diligence, deciding which companies to invest in, and portfolio management.
AI as a Part of The Decision-Making Criteria
The other side of this coin is thinking about how much to allocate to startups. What I mean by this is that there are basic platforms that help with simpler tasks like running research and reporting. One person now has the ability to become significantly more productive, without even needing technical skills by using “pick your commercial AI platform”. In my view, I wouldn’t want to allocate capital to a company that’s not using AI in some form as an operational tool either to improve productivity of employees, or to improve the company’s product, or to cut other expenses. If the business doesn’t use AI it's a serious concern because the founders or CEOs are neglecting options for potentially higher gross profits and/or lower operating expenses. So then I would ask the business to re-evaluate the amount of capital that the business is trying to raise, its allocation, as well as seriously question the competitiveness of that business in today’s environment.
Now that obviously won’t be as simple to apply for every type of business in the same way but the point is that the management tier needs to always look to get an edge by taking advantage of hyper-scaling technology.
Wrapping Up
The actual impact of the various agents that the firms above are using is still unclear — meaning the return on the amount of investment that is going into developing these agents is still unclear. What is guaranteed however, is that AI is making capital allocators more productive, and giving them the ability to have a detailed 360 degree view of any landscape before allocating capital. I can attest to this because I use AI in a very basic form for my work on a day-to-day, and my productivity has improved tremendously.
If you’re in the private capital markets, how are you using AI in your day-to-day? If you’re in any other industry, I’m also interested to know how AI agents are transforming your industry. Comment below or send me an email.
On my journey to becoming a master capital allocator, one lesson down, a billion more to go.
Hope you all have a great day
-Mansa
Share this post