Americans spend 7 billion hours every year filing taxes. It is often the most complex financial transaction a consumer does all year, in part due to a tax code that is 80,000 pages. And despite (or because of) its complexity, the tax industry is perfectly suited for artificial intelligence. The progress in tax filing artificial intelligence is a yardstick of what’s to come in the consumer finance software industry.
In response to the AI technology transformation, Intuit TurboTax has been on a journey to transform its offerings with artificial intelligence technologies to be highly personalized and require minimal user effort. Instead of “do it yourself,” it’s "do it for me.” The company has set a bold, disruptive vision, which aims to give customers what they’ve always wanted—taxes filed with minimal effort, thanks to never-enter data.
This vision is not too far off, thanks to Intuit’s Tax Knowledge Engine. Rather than a collection of screen-by-screen paths that users can take, the engine is now modeled with human logic, constraint-based and goal-directed. The engine constructs the customer’s path dynamically, given everything Intuit knows about them at any point along the way. The technology provides the foundation for intelligent systems and rules-based knowledge engines across the product, enabling features such as smart reminders and intelligent error detection, which automatically checks a tax filing to spot any omissions or errors, and also underlies ExplainWhy, which provides personalized, bite-sized explanations behind deductions, credits, and tax refunds. The Tax Knowledge Engine has reduced lines of code by 10x and helped to reduce tax filing time by up to 40%.
Gang Wang discusses Intuit’s progress so far and its vision to build a general intelligence in the field of tax filing, which includes efforts to:
Gang Wang is an Engineering Fellow at Intuit, responsible for the core tax engine that is used by 33 million US tax filers every year across mobile, web, and desktops. He has initiated and been leading the design and implementation of the next generation intelligent tax engine. His work focuses on large scale intelligent systems, knowledge engineering, enterprise architecture for mission critical financial applications. Gang is well published academically having written 26 papers, articles and book chapters. His publications cover speech recognition, natural language understanding, computer architecture, and electronic design automation, and financial software. He also holds 23 issued patents in US and Europe. Gang has a Ph.D. in Electrical and Computer Engineering from University of California at Santa Barbara.
©2017, O'Reilly Media, Inc. • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • email@example.com