Conal: NLU at Infinitus This is part of a series of blog posts featuring some of the amazing talent we have working on the technology behind Infinitus. This week, we’re excited to be featuring Conal Sathi. Conal has been building the computer brains that drive the conversations that the Infinitus digital assistant has. From rule based systems to self-learning models, Conal has been pulling out all stops to make sure our customers are getting the most out of their Infinitus experience. Why did you choose to join Infinitus? When joining a company I look for three things: mission, technology, and people. I love the mission: using AI to automate routine tasks done in healthcare. Healthcare is a space that is severely understaffed and under resourced especially during this time. Through automation, our system enables patients to get access to care faster and so our customers can be freed up to focus on higher leverage tasks. I love the technology we’re focusing on: conversational AI. The last three years have brought many critical breakthroughs in Natural Language Understanding and speech technology. We can do things with conversational AI today that simply weren’t possible a few years ago. I feel this technology can dramatically affect how things are done across multiple industries. Most importantly, I love the people in the team. I’ve known the CEO for more than a decade and really love his focus and dedication. The members in the team are all ambitious yet humble and are fun to work with on a daily basis. They all bring different skills and perspectives to the table and are eager to learn from one another. What are you working on right now? Errors in speech recognition have snowball effects on the quality of the responses of our conversational assistant. Off-the-shelf engines (such as the ones provided by Google, Amazon, Microsoft, Nuance, etc.) perform well in a lab setting when the speakers have American accented English and use common words on a studio quality microphone. Our product has to work on 8 khz phone lines and with speakers with a varied set of accents from all around the world who are using healthcare-specific terminology. I’ve been spending the last few months building a custom layer on top of the existing speech recognition systems to improve the quality of the speech recognition. Adding custom language models leverages the fact that we are dealing in a specific domain with specific vocabularies and the context within the conversation. It’s exciting to see that the work I’m doing has a direct impact on the progress of our product and on back office healthcare workers around the country. What is the favorite part about working for Infinitus? I love working on challenging problems that push me to think creatively and spend time learning and growing my skill set. We’re working on problems that haven’t been widely explored – spoken conversational AI in real-time with near-perfect precision. And of course I like working with these problems with a team that is hungry, moves quickly, and challenges me to think in different ways. What are 3 words to describe Infinitus? Optimism, Humility, and Teamwork. What advice do you have for prospective Infinitus candidates? In these times, try to find work that does good in this world and try to find people that resonate with you. What’s something you are passionate about outside of work? I’m super passionate about music, especially fusing different styles together. I’m currently studying Indian traditional music, jazz, and Western classical music on the piano (it’s amazing the power of technology to facilitate distant learning with some very accomplished teachers out there!). I recently founded a band called Raga and Blues, where we mix Indian and Western styles of music together. I feel music can be a universal language that can bring people from different cultures and backgrounds together.