Artificial intelligence has reached the inflection point where it’s less of a trend than a core ingredient across virtually every aspect of computing. These companies are applying the technology to everything from treating strokes to detecting water leaks to understanding fast-food orders. And some of them are designing the AI-ready chips that will unleash even more algorithmic innovations in the years to come.
For enabling the next generation of AI applications with its Intelligent Processing Unit AI chip
As just about every aspect of computing is being transformed by machine learning and other forms of AI, companies can throw intense algorithms at existing CPUs and GPUs. Or they can embrace Graphcore’s Intelligence Processing Unit, a next-generation processor designed for AI from the ground up. Capable of reducing the necessary number crunching for tasks such as algorithmic trading from hours to minutes, the Bristol, England, startup’s IPUs are now shipping in Dell servers and as an on-demand Microsoft Azure cloud service.
Read more about why Graphcore is one of the Most Innovative Companies of 2020.
For tutoring clients like Chase to fluency in marketing-speak
Ever tempted to click on the exciting discount offered to you in a marketing email? That might be the work of Persado, which uses AI and data science to analyze what type of email marketing call to action might work best on you. The company’s algorithms use a brand’s prior data about the ways anonymized individuals respond to its marketing emails—and it works. In 2019, Persado signed contracts with large corporations like JPMorgan Chase, which signed a five-year deal to use the company’s AI, and its revenue has grown 39% between 2016 and 2018.
For becoming a maven in discerning customer intent via messaging apps
We may be a long way from AI being able to replace a friendly and knowledgeable customer-service representative. But LivePerson’s Maven is helping companies get more out of their human reps. The machine-learning-infused service routes incoming queries to the best agent, learning as it goes so that it grows more accurate over time. It works over everything from text messaging to WhatsApp to Alexa. With Maven and LivePerson’s chat-based support, the company’s clients have seen a two-times increase in agent efficiency and a 20% boost in sales conversions compared to voice interactions.
For catalyzing care after a patient’s stroke
When a stroke victim arrives at the ER, it can sometimes be hours before they receive treatment. Viz.ai makes an artificial intelligence program that analyzes the patient’s CT scan, then organizes all the clinicians and facilities needed to provide treatment. This sets up workflows that happen simultaneously, instead of one at a time, which collapses how long it takes for someone to receive treatment and improves outcomes. Viz.ai says that its hospital customer base grew more than 1,600% in 2019.
For transforming sketches into finished images with its GauGAN technology
GauGAN, named after post-Impressionist painter Paul Gauguin, is a deep-learning model that acts like an AI paintbrush, rapidly converting text descriptions, doodles, or basic sketches into photorealistic, professional-quality images. Nvidia says art directors and concept artists from top film studios and video-game companies are already using GauGAN to prototype ideas and make rapid changes to digital scenery. Computer scientists might also use the tool to create virtual worlds used to train self-driving cars, the company says. The demo video has more than 1.6 million views on YouTube.
For bringing savvy to measuring the value of TV advertising and sponsorship
Conventional wisdom has it that precise targeting and measuring of advertising is the province of digital platforms, not older forms of media. But Hive’s AI brings digital-like precision to linear TV. Its algorithms ingest video and identify its subject matter, allowing marketers to associate their ads with relevant content—such as running a car commercial after a chase scene. Hive’s Mensio platform, offered in partnership with Bain, melds the company’s AI-generated metadata with info from 20 million households to give advertisers new insights into the audiences their messages target.
For moving processing power to the smallest devices, with its low-power chips that handle voice interactions
Semiconductor company Syntiant builds low-power processors designed to run artificial intelligence algorithms. Because the company’s chips are so small, they’re ideal for bringing more sophisticated algorithms to consumer tech devices—particularly when it comes to voice assistants. Two of Syntiant’s processors can now be used with Amazon’s Alexa Voice Service, which enables developers to more easily add the popular voice assistant to their own hardware devices without needing to access the cloud. In 2019, Syntiant raised $30 million from the likes of Amazon, Microsoft, Motorola, and Intel Capital.
For plugging leaks that waste water
Wint builds software that can help stop water leaks. That might not sound like a big problem, but in commercial buildings, Wint says that more than 25% of water is wasted, often due to undiscovered leaks. That’s why the company launched a machine-learning-based tool that can identify leaks and waste by looking for water use anomalies. Then, managers for construction sites and commercial facilities are able to shut off the water before pipes burst. In 2019, the company’s attention to water leaks helped it grow its revenue by 400%, and it has attracted attention from Fortune 100 companies, one of which reports that Wint has reduced its water consumption by 24%.
For serving restaurants an intelligent order taker across app, phone, and drive-through
If you’ve ever ordered food at a drive-through restaurant and discovered that the items you got weren’t the ones you asked for, you know that the whole affair is prone to human error. Launched in 2019, Interactions’ Guest Experience Platform (GXP) uses AI to accurately field such orders, along with ones made via phone and text. The technology is designed to unflinchingly handle complex custom orders—and yes, it can ask you if you want fries with that. Interactions has already handled 3 million orders for clients you’ve almost certainly ordered lunch from recently.
For giving birth to Kai (born from the same Stanford research as Siri), who has become a finance whiz
Kasisto makes digital assistants that know a lot about personal finance and know how to talk to human beings. Its technology, called KAI, is the AI brains behind virtual assistants offered by banks and other financial institutions to help their customers get their business done and make better decisions. Kasisto incubated at the Stanford Research Institute, and KAI branched from the same code base and research that birthed Apple’s Siri assistant. Kasisto says nearly 18 million banking customers now have access to KAI through mobile, web, or voice channels.
Read more about Fast Company’s Most Innovative Companies: