To cap off our AI-focused Transform 2019 event, taking place July 10-11 in San Francisco, VentureBeat will recognize and award emergent, compelling, and influential work in AI through the inaugural AI Innovation Awards. We take pride in shining a light on innovation through that coverage, but these awards, presented by Ople, give us a chance to do so in a new way. This is the first of a series of articles highlighting the nominees is each of our five award categories: NLP/NLU, Business Application, Computer Vision, AI for Good, and Startup Spotlight.
Natural Language Processing (NLP) and Natural Language Understanding (NLU) are increasing in importance. Smart assistants continue to find a place in our homes and offices, translation services create a more global world, and the technologies enable all manner of useful productivity tools. Use cases continue to emerge, and the capabilities of NLP and NLU continue to improve.
Denmark-based Corti describes itself as “artificial intelligence with a purpose.” That purpose is specific: To more quickly and accurately help emergency call operators determine if a call involves a cardiac event. Using pattern recognition with words and tone of voice, Corti says its significantly more accurate than human operators at identifying such emergencies, and can do it 30 seconds faster.
Rasa has its sights set on revolutionizing the field of conversational AI. Not only does the young company provide open source conversational AI that anyone can use and contribute to, it also provides a community to advance the field and organizes conferences to bring everyone together. Companies can use Rasa’s tools to make their text- and voice-based chatbots perform better with contextual conversations for applications like sales, marketing, customer service, and more.
Mozilla Common Voice
Simply put, Mozilla’s Common Voice project is designed to collect as much data as possible on what human voices actually sound like. It’s a crowdsourced means of collecting data wherein participants actively engage rather than having to opt out, which itself is notable. The end goal is to create a multilingual, open-source dataset that anyone can use to build voice-recognition into applications and services. To date, Mozilla Common Voice’s dataset comprises some 1,400 hours of voice samples across 18 languages.
The idea of “augmented writing” is what Textio is all about. The company’s technology is centered around helping organizations write better job postings by using data to score writing on a 100-point scale, offering real-time suggestions for how best to turn a phrase, reducing biased language, and giving insight into the culture that a company’s internal writing engenders.
The Transform AI Innovation award ceremony will be held on the evening of July 11 to cap off Transform 2019.