Conversation
Highlights

From Refugee to AI Pioneer: How CoactiveAI is Building the Future of Enterprise Visual Intelligence

A chance encounter over moving a couch in July 2020 sparked the creation of a company that would raise $44 million to transform how enterprises work with visual content. In a recent episode of Category Visionaries, William Gaviria Rojas, Field CTO and Co-Founder of CoactiveAI, shared how they’re pioneering the multimodal application platform category.

The parallels between two technological revolutions shaped CoactiveAI’s founding vision. William and his co-founder Cody, having witnessed the big data revolution during their time at MIT and Stanford from 2009-2013, recognized a similar inflection point approaching: “We saw a similar setup where the next generation of the next Databricks, the next Snowflake, was going to be built not just around data, but data and AI,” William explains.

The key difference? While the first wave focused on structured data in tables and logs, the next frontier would tackle the much harder challenge of unstructured content – images and videos that make up over 80% of enterprise data today.

Building Against the Current

In 2021, launching an AI company wasn’t the obvious choice it might seem today. “Roughly 40 to 50% of them, their initial gut reaction was like, ‘Hey, why are you making an AI company? This doesn’t make any sense,'” William recalls of their early investor conversations. But a16z’s Martin Casado saw beyond the skepticism to their deeper vision.

Instead of waiting for the infrastructure to catch up, they built it themselves. “One of the first things I had to do was essentially build something that acted like a vector database – not because I wanted to, but because they didn’t exist yet,” William notes. This foundation-first approach would prove crucial as the AI landscape evolved.

Finding Their Market Through Discovery

Rather than trying to serve every vertical immediately, CoactiveAI went through what they call a “market annealing” process – carefully discovering where their technology could have the most impact. Through extensive customer conversations, they found their perfect market in media and entertainment companies.

The reason? “You have visual, you have audio, and you have a temporal component connecting both – which is really rich from both perception and reasoning perspectives,” William explains. More importantly, these companies were still relying heavily on manual processes, creating an opportunity to drive massive efficiency gains.

The Power of True Partnership

When the AI hype cycle kicked into overdrive post-ChatGPT, CoactiveAI maintained their edge through deep technical differentiation and a focus on real enterprise partnerships. “If you just come to somebody like a vendor and that’s really all you’re kind of doing, this is just kind of a transactional sale,” William observes. Instead, they focused on building true partnerships by deeply understanding customer pain points.

This approach has led to successful deployments with major enterprises like NBCUniversal and Thomson Reuters. “We showed them there’s actually a different way, that there’s a way you can actually not have to do 90+ percent of the work,” William notes of their impact on customer operations.

From Prototypes to Production

As 2025 unfolds, CoactiveAI is riding a significant shift in enterprise AI adoption. “This is going to be the year where folks are switching from ‘let’s do a prototype’ to ‘let’s actually see the real things in production,'” William predicts. The focus has moved from proving technical possibilities to demonstrating concrete ROI.

This evolution plays to CoactiveAI’s strengths. While others rushed to wrap APIs around ChatGPT, they built deep technical moats by solving hard multimodal AI problems at scale. Their platform helps enterprises unlock value from their visual content through AI-powered search, tagging, and analytics – capabilities that become more crucial as unstructured data continues to grow.

Building Ladders for the Future

For William, success isn’t just about building a valuable company – it’s about creating opportunities for others. Drawing from his own journey from Colombian refugee to MIT graduate to successful founder, he’s focused on proving that success in tech isn’t limited by background.

“Let’s succeed right now and pull each other up. Then let’s build a ladder,” William emphasizes. “Let’s make sure there’s 10,000 Latino founders in the future.”

This vision of success – combining technical excellence with meaningful impact – continues to guide CoactiveAI as they build the future of enterprise AI applications. For B2B founders, their journey offers valuable lessons in category creation, market discovery, and the power of focusing on real customer problems over technical hype.