Data should increase any decision made by modern business. But most companies have a huge dead point: they don't know what is happening in their visual data.
Coactive is working on it. The company, founded by Cody Coleman '13, Meng '15 and William Gaviria Rojas '13, created an artificial intelligence platform that can understand data such as images, audio and video to unlock new observations.
The Coaction platform can immediately search, organize and analyze unstructured visual content to help companies make faster and better decisions.
“In the first revolution of large data sets, companies have become better to obtain values from their structured data,” says Coleman, referring to data from tables and spreadsheets. “But now about 80 to 90 percent of the world is unstructured. In the next chapter of large data sets, the company will have to process data such as images, video and audio on a large scale, and AI is a key element of unlocking this possibility.”
Coaktive is already working with several large media and retail companies to help them understand their visual content without relying on manual sorting and tagging. It helps them get the right content of users faster, remove explicit content from their platforms and discover how specific content affects users' behavior.
To put it more, the founders believe that Coaktive serves as an example of how AI can enable people to work more efficiently and solve new problems.
“The word Coactive also means cooperation, and this is our great vision: helping people and machines in cooperation,” says Coleman. “We think that the vision is now more important than ever, because artificial intelligence can separate us or connect us. We want Coaction to be an agent that connects us and gives people a new set of superpower.”
Giving a vision of computers
Coleman met Gaviria Rojas in the summer before the first year of the Mit Interphase Edge program. They both came to the specialization of electrical engineering and computer science and would work on the introduction With OpenCurseware Content of Mexican universities, including projects.
“It was a great example of entrepreneurship,” recalls Coleman from the OpenCursewareware project. “It was really authorized to be responsible for the activities and development of the software. This led me to the establishment of my own small internet development companies, as well as completing (MIT course).
Coleman for the first time examined the strength of artificial intelligence in the myth during work as a graduate of the researcher at the Office of Digital Learning (currently the Open Learning myth), where he used machine learning to examine how people learn at MITX, in which they organize mass, open online courses created by the myth of fabies and instructors.
“It was really amazing for me that it was possible to democratize this transformational journey through which I switched to a myth with digital learning – and that you can use AI and machine learning to create adaptation systems that not only help us understand how people learn, but also to provide more personalized educational experiences around the world,” Coleman talks about Mitx. “It was also the first time I could discover the video content and apply AI to it.”
After Mit Coleman he went to Stanford University for a doctorate, where he worked on lowering barriers in using AI. Research led him to cooperate with companies such as Pinterest and Meta in AI and machine learning applications.
“That's where I could see in the corner in the future what people wanted to do with artificial intelligence and their content,” recalls Coleman. “I saw the leading companies use artificial intelligence to drive business value, and this is where the initial spark comes for CoACTIVE.
Meanwhile, Gaviria Rojas He moved to Bay Area in 2020 and began working as Data scientist on eBay. As a move, he needed help in transporting the couch, and Coleman was a happy friend he called.
“While driving, we realized that we both saw an explosion around data and artificial intelligence,” says Gaviria Rojas. “In the myth we got a place in the first place for the Big Data revolution and we saw how people invent technologies to unlock the value from this large -scale data. Cody and I realized that we had another powder barrel to explode with companies collecting a huge amount of data, but this time it was multimodal data such as images, video, audio and text.
The platform that the founders built – what Coleman describes as “AI operating system” is an agnostic model, which means that the company can mention AI systems under the hood when the models are still improving. The Coactive platform includes pre -built applications, which business customers can use such things as searching their content, generating metadata and conducting analytical to extract insights.
“Before artificial intelligence, computers will see the world through bytes, while people saw the world through a vision,” says Coleman. “Now with the artificial intelligence of the machine can finally see the world like us, and this will cause blur of digital and physical worlds.”
Improving the human-computers interface
The Reuters painting database provides millions of photos to world journalists. Before Coaction, the company relied on reporters manually introducing markers with each photo, so that the appropriate images appear when journalists were looking for some items.
“Browsing all these raw resources was amazing slow and expensive, so people just didn't add tags,” says Coleman. “This meant that when looking for things, the results were limited, even if the appropriate photos were in the database.”
Now that journalists on the Reuters website choose “Turn on AI search”, Coaktive can raise the appropriate content based on understanding by the AI system for details on any image and video.
“It significantly improves the quality of results for reporters, which allows them to tell better, more accurate stories than ever before,” says Coleman.
Reuters is not alone in the fight to manage all his content. Digital Asset Management is a huge element of many media and retail companies, which often rely on handmade metadata to sort and search this content.
The next Coaction customer is Fandom, which is one of the world's largest platforms for information about television programs, video games and movies with over 300 million active users per month. Fandom uses Coactive to understand visual data in its internet communities and help in removing excessive blood and sexualized content.
“Fandom took from 24 to 48 hours to view every new fragment of the content,” says Coleman. “Now, thanks to COACTION, they codified their community guidelines and can generate smaller information in an average of about 500 milliseconds.”
With each case, the founders perceive Coaction as enabling a new paradigm in the way people work with machines.
“In the whole history of human-computers interactions, we had to bend over the keyboard and mouse to enter information in a way that the machines could understand,” says Coleman. “Now, for the first time, we can simply speak naturally, we can share images and videos with AI and it can understand that content. This is a fundamental change in the way we think about human-computers interactions. The basic vision of Coaction is this change, we need a new operating system and a new way of working with content and AI.”
