Apple’s Siri team lost its head of natural language understanding this month. Rushin Shah left his post as a senior machine learning manager on Apple’s virtual assistant to join Facebook’s Applied Machine Learning team, where he’ll be working on natural language and dialog understanding, according to a LinkedIn post.
Shah will be based out of Facebook’s Menlo Park headquarters, according to his profile. As the name implies, the Applied Machine Learning team is responsible for figuring out ways to implement machine intelligence in Facebook’s products.
The news comes at a pivotal time for Apple’s virtual assistant, which is getting a major update with the launch of iOS 11, which entered developer beta last week. (It’s expected out this fall.) Shah’s responsibilities at Apple included a natural language runtime system, machine learning models, model deployment pipelines, and data science for mining insights about user intents and behavior.
Earlier in his career with Siri, Shah was responsible for a number of key features, including the implementation of classifiers for determining when users are trying to have casual conversations with Apple’s assistant and a system that improved named entity recognition.
It’s unclear who will be taking over his position at Apple, at a time when Siri faces tough competition from the likes of the Google Assistant, Amazon’s Alexa, and Microsoft’s Cortana. Last week, Apple announced the HomePod, a wireless speaker that also allows users to talk with Siri through an appliance that is ensconced in their home rather than on an Apple computer or mobile device.
Natural language understanding capabilities like the ones that Shah worked on are critical to the success of a virtual assistant. It’s impossible for Siri to take action on a voice command if it doesn’t understand what someone is saying, and understanding how people speak is a very difficult problem.
Apple and Facebook did not immediately respond to requests for comment on this report.