stub Here’s What you Should Know About Evaluating an AI Startup for Investment - Unite.AI
Connect with us


Here’s What you Should Know About Evaluating an AI Startup for Investment

Updated on

By Salvatore Minetti, CEO, Fountech.Ventures

Interest in the deep tech space has been growing in recent years, particularly within the investment community. And of all the sectors operating within the realm of deep tech, artificial intelligence (AI) has become a burgeoning market to watch.

According to data from the National Venture Capital Association, $19.98 billion was raised by 1,509 AI companies in the US alone in 2019. This figure will grow in the years to comes, even if there is a short-term term dip due to the pandemic. In fact, AI start-ups promising to help us overcome the challenges posed by COVID-19 could well fuel greater investment in this field.

For venture capitalists (VCs) keen to make inroads into this space, evaluating AI start-ups for investment can be daunting. Below are some key considerations to keep in mind when seeking out the best AI talent to support.

Pinpointing true deep tech technologies

The first hurdle many investors will be confronted with is how to distinguish between genuinely innovative solutions, and those which simply masquerade as such. After all, AI is a victim of its own success – many start-ups look to bolster their commercial offering, and attractiveness in front of investors, by claiming to be “powered by AI” when, in reality, there is no sophisticated use of the technology within their core business.

Further to this issue, it is important that investors keep in mind the constraints that early-stage businesses will face as they look to establish themselves in the AI market.

Machine learning, publicly accessible libraries, pretrained models and APIs have all served to lower the barriers to entry for entrepreneurs and start-ups. Companies launching a product using these toolsets alone will likely have a myriad of competitors in no time. Naturally, this poses a risk to investors.

To mitigate this, I would urge VCs to look for start-ups that are innovating at both the science and application level. These AI companies will be inventing novel AI for their purposes and building the underlying infrastructure as they do so.

This necessarily involves separating application-level companies, which simply regurgitate third-party APIs, and those that have intense and unique research at their core. Indeed, true deep tech is novel, and represents significant advances over technologies currently in use.

Those with little prior experience in the field might be concerned about their ability to screen AI firms and determine which ones are truly pushing the frontiers of the technology. There are a number of ways around this.

In order to have early exposure to deep tech and effectively evaluate AI talent, VCs might consider building out their in-house technical tech. In effect, this would involve having a PhD on payroll to provide the appropriate technical proficiency. In doing so, investors will create the capacity to screen companies before there is even a product and market traction.

Alternatively, they might look to partners to do that for them. VCs have the option to co-invest with investors that already have in-house scientists and a solid understanding of deep tech in order to better select their investees as well as provide appropriate technical support in the early stages of their journey.

What are the traits and characteristics to look for in a founding team?

The underlying technology is a critical factor when it comes to assessing an AI start-up. Investors must be confident that a product is genuinely innovative, effectively fulfils a market need, and is commercially viable in the long-term. As part of this, the architecture behind the solution will also need to be considered to ensure it can handle increasing inputs of data and can be scaled over time.

To rest assured that all of the above points are addressed, investors should ensure all the critical roles are filled by those with proven experience and knowledge in the field. The team’s systems architects, data engineers, data scientists and DevOps engineers should all be able to demonstrate appropriate qualifications and previous field experience.

Beyond the obvious technical skills, it is important to remember that AI isn’t just about algorithms and data. It’s about people, too. For this reason, VCs should pay close attention to the traits and characteristics that founding teams display. While there’s no set criteria to follow, here are a few traits that are likely to determine the success of an AI venture.

The first is a good awareness of relative strengths and weaknesses. A founder might, for instance, have a compelling vision and the required technical knowledge to see it through. As is so often the case with fledgling companies, however, the founder might lack the appropriate business acumen to overcome common hurdles.

A high-performing AI team will be able to demonstrate a willingness to seek help and onboard the right talent to fill any existing skills gaps. A company’s culture should also be reflective of their drive to innovate: a desire to seek out critical feedback from peers, customers and experts will go a long way towards overcoming technical and business challenges that arise along the journey and help teams to focus on the big picture.

Most importantly, however, a great team will display a positive attitude: a crucial requirement for any venture in the competitive AI space. A determination to make things work, even when times are tough, will separate teams who have what it takes to scale an AI company, and those who don’t.

Salvatore Minetti is the CEO of Fountech.Ventures, which acts as venture builder and investor for deep tech and AI startups. With a presence in Austin, Texas, US, and London, UK, the company supports startups through the stages of ideation, development, commercialisation and funding.