Over the past six months, we have witnessed a sharp increase in interest in this subject and numerous attempts by various startups to automate parts of the investment process (sourcing, screening, etc.) based on their own AI algorithms or open models. Of course, the topic itself is on the cutting edge of venture capital innovations, and like any innovative technology, it still has a long journey ahead to capture the market and confirm its Product-Market Fit (PMF). However, even today it is evident that there is no other path of development, as venture risk decreases significantly when you analyze millions of unsuccessful investments and clearly understand what you’re not willing to invest in.
For me, the primary use case of AI in venture business remains the task of creating a “public company reputation” and the ability to automatically consolidate all of its public data into a unified dashboard. This dashboard would help visualize a digital snapshot of the company and indicate how actively the company is represented in the public sphere and what sentiment is primarily associated with mentions of the company and its founders. Certainly, there are many nuances here, but I am confident that such a rating can greatly simplify investment decisions, especially at an early stage, where essentially we are investing in the founders’ vision and, to some extent, growing technological trends.
The second case, for me, involves having a list of companies with which there has been past interaction and information exchange, but the investor decided not to invest and opted to wait or didn’t fully grasp the company’s business at that time. In this scenario, having a rating and understanding of the sentiment is very valuable, as the investor can automatically track the development of companies without having to gather updates or communicate with CEOs. This system would show if a company from the “non-invested portfolio” is growing better or exhibiting a more positive sentiment in the market, triggering a reconsideration of engagement and collaboration negotiations.
Lastly, the final case pertains to startups themselves that are seeking funding and are unsure which investor to approach. In this reverse situation, the startup observes a list of investors, watches how its “sentiment rating” is building up, and how it’s publicly presented. This will likely lead to a significant shift in market information availability and democratization of data about startups and investors. We at Wale.ai and The Garage see that companies are motivated to share information about themselves, but they lack the tools for continuous monitoring of this information. However, these tools will certainly emerge, and this will lead to startups being able to see how many investors are monitoring them and roughly where they stand in comparison to other startups in the portfolio, which will undoubtedly add more motivation to improve their data quality.
In conclusion, I’d like to highlight that a well-known American fund, Signalfire, is actively creating its Beacon platform based on these principles. CEO Chris Farmer describes how Beacon tracks over six million companies in real-time using ten million data sources (https://signalfire.com/tag/beacon-ai/).
Another example is Labx Ventures with their product RubX, which also helps gather data and build investment assumptions based on it.