When Rebellion Ventures raised its first fund, it focused on what its founders call “artificial Super Capability.” They intentionally sought out commitments from successful entrepreneurs, top executives in large tech companies, and other industry experts, rather than purely financial investors.
The goal was to assemble a group of LPs that were “aligned with mission and [who] are excited about working with entrepreneurs, and helping them succeed,” managing partner Jukka Alainen told Venture Capital Journal.
According to a filing, the firm recently closed on $12.9M for Rebellion Ventures I after being oversubscribed in its initial $10M target size. The majority of its LPs consist of individuals who make personal angel investments. A limited number of small funds-of-funds and family offices are also included.
Alanen said that the vast majority of over 70 individual investors were “either successful entrepreneurs, or operating executives such as [chief revenue officers], chief marketing officers, and chief technology officers], and/or industry experts in specific areas of AI and automation.” Many work in later-stage startups and some work for leading companies in the space, such as Google, Microsoft, AWS, and others, but they are not investing on behalf their companies.
Alanen, who has a background in start-up building and operating, “saw the need to get real industrial relations and operators who could accelerate the business while we upscale the businesses and tap into this expertise and these relationships.”
He said that company founders should be cautious about strategic corporate venture capital, “because it could limit your strategic flexibility.” “And traditional strategic CVC arms often invest later, maybe at Series A or even beyond, whereas our investment is in pre-seeds and seeds. The community of investors allows for greater flexibility and freedom in terms of the types of companies that we can invest in.
Rebellion, located in San Francisco, coined “artificial Super Capability,” or ASC to distinguish it from artificial general intelligence (AGI), traditionally viewed as the next step in AI to match the human general intelligence level, and artificial superintelligence (ASI), an abstract post-AGI concept.
Alanen clarified this in an email: “We are focusing on the ability to drive results and not the level of general Intelligence.” “This capability to drive results can be context-specific. For example, it depends on what is important in a particular job or objective. Further, these AI entities require more than intelligence; they must be able to interact, act and operate in their environment.
He noted that for AI systems to be economically attractive, they must be able to perform better than human-level capabilities – and not just match them. “Then, it’s more cost-effective for companies to deploy AI workers or agents to perform different tasks and [free] up human workers to do other things.”
Rebellion has so far invested in 12 companies, and plans to invest in 25 more in the next 18 months. All of these are located in North America. Checks are limited to lower six-figure amounts and start at $250,000. Alanen pointed out that this works to Rebellion’s advantage in three ways.
He explained that “it helps us to be very flexible and Agile in partnering as we don’t need to have high ownership percentages and it opens up an broader range investment opportunities rather than having a rigid target such as 10% ownership.” It allows us to work with other investors or funds without having to compete over allocations. This makes it easier for small funds to generate higher returns.
Along with Alanen, Rebellion has venture partners Ramesh Rakar, a professor and AI entrepreneur at MIT Media Lab, and Aaron Aubrecht. Aubrecht was the former chief product office at Venafi. Venafi developed a machine-identity protection platform, which Thoma Bravo acquired in 2020 for $1.15 Billion. Alanen is the only full-time employee of Rebellion.
Relationships, expertise and
Alanen explained that what attracts founders to Rebellion, is not its size, but “access to industrial relations, operating playbooks, and critical expertise” that we provide with our network of more than 70 unicorn founders, industry experts and operating executives who are all LPs of our fund and members of our advisory network. Members of this community can serve on the boards of portfolio companies or provide advice in other capacities.
Alanen was previously the head of strategy, corporate, and business development for PagerDuty. He helped the company reach its $2.8 billion initial IPO in 2019 and led efforts to revamp management of its digital operations using AI and automated workflows.
He said that helping PagerDuty clients automate inefficient and slow manual processes with AI assistance gave him a “hands on understanding of AI automation and how it can deliver value to customers. It also provided insights into the landscape and tech stack, as we integrated with partners in general, and helped us build relationships and expand our networks in the industry.”
Quandri.io is a Rebellion portfolio company that focuses on helping agencies and insurance brokers replace manual processes which take up a lot human time and effort. Quandri’s AI enabled software robots manage these tasks autonomously. This reduces cost for clients up to 80 percent, as they complete data-intensive tasks 10 times faster than normal.
Alanen explained that unlike large language models or foundation models, action-oriented AI can handle more complex workflows. They are able to “execute these complex processes that have many, many steps, instead of a single task you would ask ChatGPT to perform, for example.” “And they integrate into a variety of systems and environments, just like a person would,” Alanen said.
While foundation models are important for more advanced AI agents, they do not have the ability to observe and operate independently, integrate with multiple systems, or complete more complex tasks as humans can.
Aptivio is another portfolio company that uses AI agents to drive growth in the sales pipeline for its customers. Its digital agents monitor over one billion external datapoints per day to identify prospects that could be attractive customers for a company. They can then prioritize promising opportunities, and can actively drive outreach to these customers.
Alanen said that customers can integrate Aptivio’s platform in just a few hours and be interacting with new sales leads in 24 hours. Many traditional AI go to market projects require a customized system, and long learning times. With Aptivio “you can see the results of driving additional leads and growing your pipeline within a month.”
According to industry research, AI agents that are designed for specific tasks tend to have higher success rates and quicker times to value compared to traditional AI projects. The vast majority of these projects fail, he said.
A Rebellion portfolio firm that is still in stealth is training AI agents as autonomous AI reliability engineers. This will help customers avoid downtime for their websites and applications. The agents “continuously monitor the companies’ technical environments 24/7 by integrating a number different sources and then, if anything happens, they can take autonomous action” to fix reliability problems, said Alanen.
AI agents are mainly limited to the digital realm for now. “But this phenomenon is growing and becoming more prevalent in the physical universe, so there will be [applications] that mix the digital and physical worlds,” he said.
Alanen cited the recent test of the US Air Force, which used AI-controlled fighter jets to perform a dogfight operation autonomously in the sky. In the future, industries like logistics and transportation will adopt autonomous systems in order to perform a wide range of tasks on the physical world.