Highlights
1:45 Sage’s Background and Engagement With Technology and Artificial Intelligence
9:05 How He Is Employing AI, NLP, and Real Time Continuous Learning at Agolo
16:11 Agolo’s Main Customer Base and the Work They Do for Them
20:37 Areas of Success and Challenge for AI Now and in the Future
27:23 Sage’s Founder Role Model
Sage's Insights
“An area that really differentiates us is microservices. We use micro models, where we can actually deploy our AI models into containerized environments into a private area that can then learn from the client data. And then create those knowledge graphs, create those summaries, create those co-reference resolutions from their data with minimal human eyes on it.”
“The real time continuous learning aspect has been a real breakthrough for us, because some of our initial clients have the highest editorial standards […]. To be able to read the content and then put their brand on it is a huge step for them.”
“We’ve seen empirically that the best sort of solutions, at least in the present state, are machine guided humans, not machine end-to-end processes, where you can actually get better information, more actionable information from machines, and then ultimately drive a human decision.”
“Human decisions can help drive the machine to find more related content and really continuously learn from that process. That sort of synergistic relationship is the ideal.”
Sage's Bio
Sage Wohns is the cofounder and CEO of Agolo, a service that uses artificial intelligence to summarize information faster, and with better coverage, than any person in the world. Sage believes that if we can responsibly apply the capabilities of AI, it can empower us toward insight and ingenuity of which only the human mind is capable.
After completing his MBA at Columbia Business School and leading Multilingual Engineering at Rakuten in Tokyo, Sage enjoyed a brief stint farming radish in the countryside of Hokkaido before returning home to pioneer the AI-startup boom in New York City. He blended his programming and business-strategy experience to launch Agolo, which has since become a trailblazer for the mainstream use of AI in finance and news organizations, and the winner of Citi’s Top Business Intelligence Tool prize.
Besides being a proud nerd and Seattle native, Sage is also an accomplished cellist, fluent Japanese speaker, and avid mountaineer.
Show Notes
Co-Founder and CEO of Agolo, Sage Wohns, joins Melody on the podcast today to discuss his programming and business strategy experience and how he has combined them to advance the role of AI particularly in finance and news organizations. He also reveals the keys to the founder role in which he has excelled, and the role models whose example he feels are worthy of emulating.
TECHNOLOGY AND AI EXPERIENCE
Sage’s work in AI has gone from simply trying to define it to employing NLP to help clients in an effective manner, and how to understand the vast amount of big data now available to them.
AGOLO’S AI, NLP, AND CONTINUOUS LEARNING: SCAN, ORGANIZE, SUMMARIZE
Agolo’s work is built upon a foundation of information search and retrieval, employing at least a dozen machine learning and natural language processes under the headings of scan, organize and summarize. They have become early adopters of micro models that can learn from client data – models which Sage feels represent the future of this work.
THE ADVANTAGES THAT AGOLO PROVIDES TO ITS CLIENT BASE
The areas of finance and news organizations increasingly make up the bulk of Agolo’s client base. Creating two summaries from the same document dynamically is increasingly important and valuable to these clients, and Agolo is making great strides in reinterpreting information in a simpler way to make it more accessible to a wider audience.
AI SUCCESSES AND CHALLENGES
The best sorts of solutions are ‘machine guided by humans’, not ‘machine end-to-end’, and the goal should be to understand how machines work so that we can work better and be enhanced by them. Accuracy in AI reading of documents remains the most fundamental concern, and the detection of language nuances such as sarcasm continues to be one of the most difficult problems in this area.