Delivering Context to Content

We have exciting news: Coactive AI is officially going to market with $14 million in combined seed and series A funding from Andreessen Horowitz and Bessemer Venture Partners, alongside angel investors SV Angel, AME Cloud Ventures, Fei-Fei Li, Jure Leskovec, Eli Collins, Cloud All Star Fund, Next Play Capital, and Exceptional Capital.

We've been behind the scenes for the past 18 months, building and iterating on a solution that brings context to visual content at a crucial time. Today, I get to share more about that work and what's to come.

Data overload. We've seen an explosion in visual content over the last decade. Video accounts for over 80% of traffic on the internet, up a whopping 73% since 2017. And we all interact with visual content daily, whether scrolling through Instagram, looking up a how-to tutorial, clicking on a Facebook ad for a new gadget, or snapping a picture to send to a friend. Some of these actions result in data. Now imagine that volume of visual content data (from millions of users and customers) living in organizations' systems or spreadsheets somewhere. That's hundreds of millions, if not billions, of files and data points without context.

AI is essential to activate this vast amount of unstructured visual content.

We've all seen the headlines generative AI tools are making. However, there's another parallel revolution happening in the same realm: the rise of intelligent search. 'Intelligent search' goes beyond simple search terms and understands what the inquirer is looking for on a deeper level thanks to the power of machine learning.

When it comes to image and video data, creating effective search tools has been difficult. Unlike other data types, visual data doesn't have easy-to-search identifiers – the context and meaning are in the visual itself. Without complete and thorough metadata, it's nearly impossible to process, let alone understand, the value behind image and video data.

“Labeling visual content is incredibly resource-intensive, taking away precious time that these teams could use for new innovations and discoveries. Coactive not only eliminates that problem but supercharges these teams by making it easy to search and analyze visual content. This is an immense value-add for any company that deals with vast quantities of unstructured image and video data.”

― Elliott Robinson, Partner at Bessemer Venture Partners

Tagging this type of content's metadata is a significant pain point for data practitioners, taking days or even months to wrangle. As a result, companies can't efficiently act on their data without understanding it. This inability can lead to disadvantages in capturing user behavior and even potentially harmful gaps in moderating negative content.  

This is what we're solving. When I met Will, Coactive's co-founder, we connected on our shared experience with the metadata problem and issues in data pipelines. We combined our collective expertise in academics and at enterprise tech companies to build a solution that brings intelligent search to enterprise-level unstructured image and video data.

“Coactive is solving for a very real need in analytics. Unstructured image and video data have sat practically unusable. By tapping into the underlying value of this data, companies can leverage user-generated content for sales and marketing functions, drive policies around trust and safety, and efficiently moderate content, among other use cases. Coactive makes this possible.”

― Martin Casado, General Partner at Andreessen Horowitz

Coactive processes and gleans insights from companies' visual content, with no labels or Ph.D. required, integrating seamlessly with existing data lakes and big data tools using SQL. The result is companies access a much more complete and deeper understanding of their image and video data, which informs crucial business functions. For example, the insights generated using Coactive help companies with large quantities of user-generated content understand trends in user behavior, create trust and safety policies, effectively moderate content, and more. Today, with Coactive, companies can:

• Generate metadata for large quantities of content in minutes.
• Query unstructured visual data using SQL and existing big data tools.
• Power critical systems, such as search analytics.

Our passion for applying ethical and sustainable AI to big problems is the foundation of our work at Coactive. Data has always fascinated me; it's why I spent over a decade pursuing in-depth academic studies, getting a Ph.D. from Stanford, and regularly publishing scholarly articles on these issues (most recently, I participated in a study about combating algorithmic bias in computer vision). As a founder, I set out to bridge the gap between what I had spent years studying and what data practitioners need to actually get value from the data they encounter.

Founding Coactive, Will and I knew we could create tangible progress. We're proud to work alongside a fantastic team of scientists and software engineers with decades of collective experience in data and machine learning at eBay, Google, Meta, Stanford, and other leading organizations. Together, we've built a platform and company that will change the trajectory of AI-driven data analytics.

This is only the beginning. We’ve seen how immediately impactful our technology is in our closed beta with companies like Fandom — and we’re just getting started. As companies better understand and create salient content strategies, there will be a major boost in efficiency, innovation, and, ultimately, a better experience for end users. If you’re interested in trying out Coactive for your organization, request a demo.

“Fandom recognizes the importance of understanding visual content to maintain a safe community and drive business initiatives. However, doing so has traditionally required substantial investments in time and resources. To address this challenge, Fandom has partnered with Coactive AI. Coactive AI's solution provides Fandom with a faster and more cost-effective way to view and analyze visual data, while also allowing for customization based on Fandom's specific definitions and protocols.”

― Florent Blachot, VP of Data Science & Engineering at Fandom

We want to extend a heartfelt thank you to Elliott Robinson, Martin Casado, all our investors and advisors, partners, team members, customers, family, and friends for their trust and belief in us. All of this wouldn't be possible without you, and we're excited to move into this next chapter with your support.

Want to learn more about how Coactive can help your organization leverage image and video data? Request a demo today, or reach out to us at info@coactive.ai.