Who doesn’t enjoy a good challenge? Katabat CTO Ye Zhang decided to give his artificial intelligence team the opportunity to flex their muscles in a Google Kaggle competition in December. “We are not the ones who invented machine learning (ML),” Ye recognized. “However, we have unique advantages when it comes to applying natural language processing and recommender systems in the financial industry.” Entering a recommendation system challenge, sponsored by the Association for Computing Machinery’s International Conference on Web Search and Data Mining, was the perfect chance to have fun while at the same time hone Katabat AI technologies on sophisticated datasets.
The team’s solutions placed in the upper quintile among more than a thousand teams. Ye reported that it was the continuing learning experience that felt most valuable (besides the thrill of competition!). “We’re adopting a researcher’s mindset in this kind of competition, aiming at becoming the best of the best,” he reflected. “We kept working on our ML model, even after the competition had ended, by reaching out internationally to the first place and third place teams.” The result was a solution that replicated the winning accuracy rate.
The Katabat team is very enthusiastic about incorporating ML technology in production releases in 2018. The product/sales team is reaching out to current clients to better understand their priorities for which offerings become available first.
What feature of your business would you like to see Katabat transform with artificial intelligence? Reach out to us in the comments or contact firstname.lastname@example.org.