Unitary, an AI-driven visual moderation company headquartered in London, unveiled that it has successfully raised $15 million (equivalent to €14.31 million) in its Series A funding round.
In a Series A funding round, led by Creandum and joined by Plural and Paladin Capital Group, Unitary, the London-based AI-powered visual moderation company, has secured vital investment.
Sasha Haco, Co-Founder and CEO of Unitary, emphasizes the company’s commitment to utilizing AI for a safer online experience while keeping pace with the ever-evolving landscape of internet content.
Founded by Sasha Haco and James Thewlis, Unitary has harnessed AI technology that mimics human cognition, comprehending videos and images by deciphering multiple signals to gauge content and context.
Sasha Haco further notes, “This funding empowers us to maintain our position as pioneers in video understanding research and continue our mission of enhancing the online environment.”
Gemma Bloemen, Principal at Creandum and a board member at Unitary, expresses enthusiasm for supporting the company in the crucial field of content safety, recognizing Unitary as early leaders in this space.
This funding coincides with Unitary’s expansion into multiple languages, doubling its team size, and achieving a threefold increase in daily video classification—from 2 to 6 million. The capital infusion will bolster research and development efforts, further team expansion, and the establishment of strategic partnerships with leading social platforms and brand safety organizations.
Advertisement
Active Content Moderation
The surge in video content complexity, comprising a staggering 80% of internet traffic, is projected to increase tenfold between 2020 and 2025, presenting a formidable challenge that exceeds the capacity of human reviewers.
Christopher Steed, Chief Investment Officer of Paladin Capital Group, underscores the pressing need for a technology-driven approach to detect and address harmful content in the online realm.
Unitary’s machine learning solution is adept at blending visual, auditory, and textual content analysis, offering a robust solution for platforms grappling with the complexities of online content moderation. This capability proves indispensable for platforms adapting to evolving regulations, such as the UK’s Online Safety Bill and the European Union’s Digital Services Act, which mandate more proactive content moderation.
Remarkably, the company has already achieved “seven figures” in annual recurring revenue, as reported by Ian Hogarth, Partner at Plural and a board member of Unitary. This significant milestone has prompted swift follow-up funding rounds, following a successful raise of $8 million (equivalent to €7.63 million) in March.
Ian Hogarth expresses confidence in the startup’s capacity, stating, “From the outset, Unitary possessed some of the most potent AI for classifying harmful content. We firmly believe that this team is poised to redefine how we ensure the safety of visual content in the digital age.”
Harnessing Multimodal Models
Unitary AI’s breakthrough resides in its utilization of multimodal models. Although multimodal AI research has been in progress for years, it is now finding more practical applications, and Unitary stands at the intersection of advanced research and tangible applications within this evolving field.
Sasha Haco, in a recent interview, elaborates, “Rather than simply analyzing a sequence of frames, to discern nuances like whether a video is, for instance, artistic or violent, it’s imperative to replicate how a human moderator perceives the video. We achieve this by scrutinizing text, sound, and visuals.”
In contrast to existing tools that handle data types individually, Unitary’s approach mirrors the way a human moderator views videos, reducing false positives and enhancing accuracy. Customers have the flexibility to tailor moderation parameters and frequently deploy Unitary in conjunction with human teams, lightening the workload of moderators.
While visual-only models have proven effective to some extent, the advent of multimodal moderation presents a promising growth opportunity in the market, addressing ongoing content moderation challenges prevalent in social platforms, gaming companies, and digital channels.
