Reporting by Helen Popkin, Aayushi Pratap and Nina Wolpow
The Covid-19 pandemic was devastating for many industries, but it only accelerated the use of artificial intelligence across the U.S. economy. Amid the crisis, companies scrambled to create new services for remote workers and students, beef up online shopping and dining options, make customer call centers more efficient and speed development of important new drugs.
Even as applications of machine learning and perception platforms become commonplace, a thick layer of hype and fuzzy jargon clings to AI-enabled software.That makes it tough to identify the most compelling companies in the spaceespecially those finding new ways to use AI that create value by making humans more efficient, not redundant.
With this in mind, Forbes has partnered with venture firms Sequoia Capital and Meritech Capital to create our third annual AI 50, a list of private, promising North American companies that are using artificial intelligence in ways that are fundamental to their operations. To be considered, businesses must be privately-held and utilizing machine learning (where systems learn from data to improve on tasks), natural language processing (which enables programs to understand written or spoken language) or computer vision (which relates to how machines see). AI companies incubated at, largely funded through or acquired by large tech, manufacturing or industrial firms arent eligible for consideration.
Our list was compiled through a submission process open to any AI company in the U.S. and Canada. The application asked companies to provide details on their technology, business model, customers and financials like funding, valuation and revenue history (companies had the option to submit information confidentially, to encourage greater transparency). Forbes received several hundred entries, of which nearly 400 qualified for consideration. From there, our data partners applied an algorithm to identify 100 companies with the highest quantitative scoresand that also made diversity a priority. Next, a panel of expert AI judges evaluated the finalists to find the 50 most compelling companies (they were precluded from judging companies in which they have a vested interest).
Among trends this year are what Sequoia Capitals Konstantine Buhler calls AI workbench companiesbuilding of platforms tailored to different enterprises, including Dataiku, DataRobot Domino Data and Databricks. Healthcare and biotech research, as conducted by Komodo Health, Genesis Therapeutics and Verge Genomics, remains a key area for advanced AI, as is computer vision, with companies such as Viz.ai and AMP Robotics using the technology to improve health care and waste recycling. Companies reliant on natural language processing, such as Duolingo, Lilt and Whisper, which developed an AI-enabled hearing aid, are yet another core category. Autonomous vehicles are again represented on the list, this year by Gatik, which sees middle mile driverless deliveries as a lucrative, early market to target.
Looking ahead, judge Andrew Ng, founder of Google Brain, cofounder of Coursera and founder and CEO of Landing AI, sees more opportunities for AI to help manufacturers and healthcare providers with data tailored to their specific needs.
.
.
There are plenty of open source models that you can download that work just fine for a problem, but what really needs to be customized is the data, he says. I'm finding that for multiple companies, by starting to help enterprises efficiently get the data they need to feed into an open source model, that's the key to unlocking the value for that business.
This 2021 list features 31 companies appearing for the first time, while seven have qualified for three years in a row. In terms of valuation, at least 13 of the AI 50 are valued at $100 million or less, while 13 are unicorns valued at $1 billion or more. Silicon Valley remains the hub for AI startups, with 37 of 50 honorees coming from the San Francisco Bay Area.
This years judges included: Tonya Custis, director of AI research, Autodesk; Michael Jordan, professor of computer science, University of California, Berkeley; Xuezhao Lan, founding and managing partner, Basis Set Ventures; Andrew Ng, cofounder, Coursera; computer science adjunct faculty, Stanford University; Fay Cobb Payton, professor of information technology and analytics, North Carolina State University; Gill Pratt, CEO, Toyota Research Institute; chief scientist, Toyota Motor Corp.; Carol Reiley, AI entrepreneur and scientist; former president, Drive.AI; and Raquel Urtasun, professor of computer science, University of Toronto.
Forbes
(Honorees are listed alphabetically. An asterisk donates valuation data from Pitchbook rather than company sources or Forbes estimates.)
.
Instead of viruses, spam or ransomware, Abnormal Securitys cybersecurity software targets business email compromise (BEC) attacks, which cost businesses nearly $1.9 billion in 2020, according to the FBI. Thats more than half the $3.5 billion total loses to cybercrime. BEC occurs when a bad actor compromises legitimate corporate email accounts, masquerading as an employee and tricking anyone from the CEO and CFO to the human resources manager into transferring large sums of money or sensitive documents. A lot of the conventional email security companies built technologies that stopped attacks theyve seen before by memorizing known bad behavior, says CEO Evan Reiser. But the schemes keep evolving. To combat the fast-proliferating BEC attacks, Abnormal Security instead borrows from adtech with behavioral profilingthe way Facebook and Twitter can display ads tailored to youto predict the legitimacy of emails seemingly sent by a trusted party, yet requesting money transfers of thousands or even millions of dollars.
.
AMPs robots can automatically separate plastic cans from cardboard, batteries from wires and wood from concrete. The company caters to recycling centers by offering robotics software and hardware that autonomously identifies and sorts recyclable materials. Its the brainchild of CEO Matanya Horowitz, who worked on robotic graspingteaching robots how to pick up different objectsas a Ph.D. student at Caltech. Business spiked amid the pandemic, culminating at the end of last year in a deal with Waste Connections, the third-largest waste management company in the United States, to install 25 robots across its facilities, as well as a $55 million funding round.
.
AI models in research mode can sometimes react quite differently when analyzing real-world data. But the reason why often isnt available in real time. Founded by Dhinakaran, formerly a key engineer at Uber, and Lopatecki, founder of TubeMogul, an ad-bidding platform, Arize AI is a real-time analytics platform designed to watch, troubleshoot and provide guardrails on deployed AI. In the simplest old school companies we were seeing deployments of hundreds to thousands of models, each touching customers daily, says Lopatecki. I simply couldnt see an AI future that didnt have software that provided observability around ML systems, to help troubleshoot and improve the most complex systems ever deployed.
.
Atomwise is powered by a drug discovery algorithm that uses a neural network to simulate how different small molecules may bind to a protein. The machine learning technique allows scientists to rapidly simulate the interactions of millions of molecules to determine which have potential in preclinical trials. Because of their convoluted structures and the complex ways they interact, thousands of proteins remain unlinked to any drug treatments. Atomwises 250 customers, which include research institutions like Columbia University and pharmaceutical companies like Bayer, are running nearly 800 projects across areas including cancer, clotting disorders and brain diseases. Its already helped to discover promising drugs for multiple sclerosis and ebola which were successful in animal trials.
.
Tracking ocean-bound cargo shipments is a tricky business. Rough seas, wind and weather can wreak havoc on scheduled arrivals in port and boost fuel use. Bearing wants to smooth things out by using AI to help shippers manage and track their fleets, reduce fuel consumption and optimize routes. Bearings AI models to predict vessel performance are significantly more accurate than the traditional physics-based models used in the industry, says CEO Dylan Keil. Successfully demonstrating Bearings optimization tools in theory, Keil recalls, it was still thrilling to actually guide that first 650-plus-foot vessel across the Pacific. Bearing has partnerships with major global shippers K Line and MOL.
.
For CEO Kevin Albert, the challenge is bringing robots out of the factory, as well as science fiction, and creating machines that work in the real world. A Boston Dynamics alumnus, Albert previously worked on BigDog, the four-legged military robot funded by DARPA. That experience helped him identify another massive real-world application: the ever-shifting nature of construction sites, particularly streamlining the laborious process of drywall installation. Using Canvas robots, construction workers can reduce drywall finishing times from seven to two days, while achieving an extra smooth finish. Its AI-driven technology has only been applied to construction sites in Northern California, and the startup plans to use its just-raised $24 million Series B to expand into new markets.
.
Billions of dollars are spent on customer resource management systems, yet salesarguably the most important function of any organizationremains the least efficient function of many organizations, notes CEO Andy Byrne. The reason is obvious, he says. Data entry limits time for selling, so sales teams avoid it, resulting in bad data. Clari addresses this pain point by using AI to streamline CRM updates, alleviating the data entry load from the sales team, while managing sales and forecasting with predictive insights.
.
Cofounded and chaired by Sebastian Thrun, the creator of Googles self-driving program, Crestas goal is to change the way people have conversations in the customer service industry. It uses AI to learn the most effective replies to customer questions from the best agents in a team. It then provides real-time prompts to less effective call center agents, on what could be best, or most effective, replies to customer questions. CEO Zayd Enama Stanford AI lab Ph.D. dropout who immigrated to the U.S from Karachi, Pakistan at 17says the companys approach helps those employees who are "caught in the middle" to highlight their strengths and become invaluable to their businesses going forward.
.
CrowdAI specializes in extracting meaningful information from the flood of visual data created by everything from cell phone cameras to satellites. It does this with a software platform designed to be easily accessible to all users, not just data scientists and developers. CrowdAIs technology is being used by manufacturers, the California Air National Guard and Californias Department of Forestry and Fire Protection, for which it built a custom computer vision model able to detect wildfires in nearly real time. When it comes to computer vision, we believe humans need to remain in the loop, says Devaki Raj.
.
Databricks is backed by the four cloud titansAmazon, Google, Microsoft and Salesforcein addition to blue-chip investors such as Andreessen Horowitz and Coatue. Built on top of the analytics engine Apache Spark (also created by its founders), the UC Berkeley-bred company combines the raw data repositories, or data lakes with the structured information of data warehouses to create what CEO Ali Ghodsi calls a lakehouse where companies store and make use of their data. Were able to consolidate all of a customers data workloads, across both analytics and AI, on a single platform, he says. Comcast, Credit Suisse and T-Mobile are among 5,000 customers using Databricks to build business analytics or machine learning tools. Theyve helped the company reach $425 million in annual recurring revenue; following a $1 billion funding round earlier this year, an IPO could soon be on the horizon.
.
Founded by four Frenchmen in Paris, Dataiku has expanded into a global operation that makes software for big companies like Pfizer, Sephora and Unilever that want to develop AI by themselves, but lack the resources of Amazon or Google to do so. Levis, for example, used Dataikus tools to create a machine learning-based recommendation system for its customers. After achieving unicorn status at the end of 2019 with a cash injection from Alphabet investment arm CapitalG, the company raised an additional $100 million last year.
.
DataRobot helps customerswhether seasoned data experts or coding novicesbuild their own machine learning predictive models. United Airlines, for example, used DataRobots software to predict which passengers might gate-check their bags, while the NBAs Philadelphia 76ers used its tools to help model estimates for season-ticket renewals. During the pandemic, DataRobot partnered with the federal government to identify and resolve gaps in Covid-19 information to provide visibility on hospital data such as ICU and ventilator supplies and bed shortages. We also overhauled and improved forecasting models being used by decision-makers in vaccine clinical trials to accelerate the approval timelines, says CEO Dan Wright, who took over the top role from Jeremy Achin in March.
.
As a dark matter Ph.D. student, Scott Stephenson started recording audio as part of his research, noting the silence and background noise that existed between acoustic information he meant to capture. Developing a tool that could mine that meaningful sound resulted in Deepgram, automatic speech recognition software that facilitates better transcriptions of recorded audio like that of an in-person meeting or a Zoom-based conference call. Though it is one of many companies optimizing transcriptions, Deepgram claims to be the only platform that learns based on phonetic patterns and phrases, including industry-specific terms often misunderstood by speech-to-text services.
.
The race to commercialize self-driving cars depends on high-definition maps for vehicles to know where they are, and many of the companies developing that technology were diverting resources to build their own maps. DeepMap was founded to help companies avoid that redundant effort and save money by creating a map engine as a service. It uses deep learning to automatically detect and create 3D map features and landmarks, such as street signs, signals and lane lines, from input sensor data. The technology can support millions of cars while keeping map quality high, map consumption highly efficient and cost very low, say founders James Wu and Mark Wheeler.
.
Elprin, Yang and Granade came up with the idea for Dominosoftware that gives data scientists the tools they need to create, test, and run their own AI modelswhile leading hedge fund Bridgewater Associates research organization. They understood that in order to catch up with Bridgewater and other model-driven businesses like Amazon, Netflix and Tencent, novice and veteran companies alike would benefit from access to new technologies and platforms that would enable data science as a core capability: thats what Domino aims to do. Over the coming decade, winners in all industries will be the ones that put models at their heart of their businesses, Elprin says. So far, big-name customers include Johnson & Johnson, Lockheed Martin, Dell and Allstate.
.
No matter who you are, where you come from or what you do, youve probably heard of Duolingo: the AI-driven language-learning app may already be sending daily reminders to practice your French. This kind of ubiquity is what CEO Luis von Ahnwho was a 2006 MacArthur Fellow and previously launched reCaptchais looking for. Duolingo is used by people all across the socioeconomic spectrum, from Syrian refugees to billionaires and celebrities like Bill Gates and Joe Jonas, von Ahn says. Such broad adoption, fueled partially by pandemic-boredom, led to a $750 million surge in valuation over a seven-month period in 2020.
.
A late cancer diagnosis means a higher chance of death. Using AI and automation, Ezra studies MRI scans to help radiologists detect cancer lesions faster and better. Its technology, which was approved by the FDA in October 2020, automates interpretation of the size and boundaries of cancerous lesions, faster than the amount of time a radiologist would take. CEO Emi Gal, who is at high risk for melanoma, dreams of making a $500 full body MRI for cancer, in the next 3 years.
.
Data from industrial and manufacturing operations is growing at a rate that equals human data but companies are only able to utilize about 2%, according to Falkonry CEO Nikunj Mehta. He saw an opportunity to utilize that underexploited data using AI and machine learning to help companies improve all aspects of their operations. The software platform it developed to do this is particularly suited for manufacturing, defense and energy fleet operations, says Mehta. It monitors operations and detects and predicts failures, and is already being used by customers including the U.S. Air Force, IMA Life and steel producer Ternium.
.
FarmWise AI-driven precision weeding machines help farmers streamline the process of growing high-value commodity vegetables. Serving farms in California and Arizona, Farmwise offers its technology as a service, charging a per-acre fee to weed fields rather than selling its equipment. The more times a machine visit a given farm, the more it learns, and the better at weeding it becomes. The company is also developing a grower dashboard that will allow farmers to track metrics like precise crop count, size and spacing trends within given fields.
.
As engineering manager of Facebooks News Feed team, Gade was tasked with building Why Am I Seeing This, an internal tool to help understand how its algorithm elevated certain stories. Every company should understand the inner workings of its AI models, Gade thought, and Fiddler Labs was born. We cannot allow algorithms to operate with a lack of transparency, he says. We need accountability to build trust between humans and AI. Fiddler Labs platform enables companies to analyze and understand the AI in their system while meeting regulatory compliance and building trust in the end user.
.
Cofounded by Deon Nicholas and Sami Ghoche in 2017, Forethought is an enterprise search company that created a question-answering retrieval AI agent called Agatha. It embeds into existing employee workflows, helping them work more efficiently instead of replacing them to improve customer service. Agatha solves and assists using machine learning and natural language processing that continues to improve over time. While some tasks will be automated in the future with AI, some things will also work better with a compassionate human completing the task, says Nicholas.
.
Most companies in the multibillion-dollar race to commercialize autonomous driving have focused on the toughest challenges, such as robotaxis and self-driving big rigs. Gatik zeroed in on a more near-term application: middle mile routes hauling goods on fixed, repeatable circuits. B2B short-haul logistics was where we felt we could add the most value, bring a product to market quickly and scale safely, says CEO Gautam Narang. We figured out a business model with strong economics at the same time as solving the technology for commercial deliveries. The company expects revenue to grow by up to five times in 2021 and counts Walmart as a major customer, and is developing medium-duty autonomous delivery trucks with Isuzu.
.
Genesis is on a path to discover new drugs using AI. Its software studies the chemistry of new molecules to predict which could be safe and efficacious drug options for human diseases. CEO Evan Feinberg, who earned a Ph.D. from Stanford University, says drug discovery is akin to finding a needle in a haystack. While traditional methods were error-prone and slow at predicting promising drug candidates, technological advances are helping change that, he says. Rather than applying AI solutions for image recognition or language processing to the pharmaceutical industry, Feinberg and chief technology officer Ben Sklaroff created new AI tools specifically for chemistry.
.
Thousands of sales teams rely on Gongs natural language processing capabilities to close deals and shorten sales cycles. 99% of the information shared by customers never makes it to the CRM and the 1% that does is heavily filtered, says CEO Amit Bendov of the software landscape when he started the company. Gong, he says, translates the information into higher-order insights. First, it transcribes customer emails, phone and video calls, then it employs machine learning to analyze everything from when a customer is ready to be pitched for a product refresh to which deals are at risk of being lost. When we started the company, AI/ML just crossed a threshold of being good enough for our needs, Bendov says. Had we started the company a couple of years earlier we might have not been successful.
.
Gretel is an open-source, synthetic machine learning library that helps developers anonymize the user date they need to build new and better features. Red tape and manual redaction of personally identifiable information slows the process of acquiring the workable data needed to test new ideas in the digital realm. Through Gretel, developers can generate synthetic datasets that are statistically equivalent to the sensitive information theyre based on, yet cant be traced back to the individuals within the original data.
.
Hyperscience was created to make data entry less boring, keep it from being relegated to a back-office chore and make it a seamless part of a companys everyday operations, regardless of the strength of that organizations machine learning team. Rather than take this work out of the hands of human workers, the company develops collaborative, AI-driven solutions that adapt to unique data-entry problems. Weve created a new class of software that deliberately divides work between people and machines based on the needs of the task, CEO Peter Brodsky says. The team attributes a 10-fold increase in platform usage over the last year, leading to a doubled employee base, to the appeal of this approach.
.
Icertis software takes contracts and extracts the data in them. Trained on 10 million contracts across more than 40 languages, its artificial intelligence puts an eye on every contract for use cases from simple automation of administrative tasks to analyzing risks or ensuring compliance. About 225 customers, including Apple, Johnson & Johnson, Porsche and Microsoftwhere CEO Samir Bodas was previously a directoruse Icertis, which in March raised $80 million in a Series F funding round.
.
When the government introduced the Paycheck Protection Program, banks had to scramble to process unstructured data like scanned W-2s and pay stubs. Instabase, a platform that allows businesses to build customizable apps to automate different parts of their businesses, tackled this new pain point. For seven days, the company worked around the clock creating an app that allowed banks to process hundreds of thousands of PPP loan applications a day. It was scary, and there were uncertainties and anxiety,'' says CEO Anant Bhardwaj. With products that help customers integrate third-party models, the company hopes to provide similarly effective solutions for customers in every sector.
.
Founded by a pair of longtime healthcare consultants, Intelligencia bucks the trend of pharmaceutical companies searching for drugs with their own AI. Instead, it partners with existing pharmaceutical companies to provide software thats meant to minimize the risk of failure in drug development and clinical trials. The startup uses AI to predict the likelihood that a clinical trial will succeed and also provides input on how to improve the trial or what other areas of research to target. Our strong belief is that biotech needs to catch up to baseball and its own Moneyball moment is here, says cofounder Vangelis Vergetis, referencing the 2011 film in which a small-budget baseball team used advanced analytics to outperform expectations.
.
Contracts are the atomic unit of modern business, says CEO Jason Boehmig, who had the idea to bring intelligent software into the legal world while he was still in law school at Notre Dame. After leaving his first law firm job, Boehmig, along with chief technology officer Cai GoGwilt, launched Ironclad, which uses AI to digitize contracts and related processes, creating useful data packages. Ironclads AI capabilities are built in partnership with Google Cloud AI and allow customers to upload legacy contracts 40% to 50% faster than they could before, according to the startup. The company raised a $100 million Series D in January, which it plans to use to drive product innovation and scale go-to-market functions.
.
The backbone of Komodo Health is a map which compiles the clinical encounters of 325 million patients who go through the healthcare system. The end result is a massive web of data that allows Komodos more than 100 customerswhich span government agencies, healthcare payers and pharmaceutical firmsto uncover a slew of clinical and business insights. Komodos analytics-driven AI algorithms allow for use cases that include forecasting the market for a drug, identifying potential patients for a clinical trial and tracking the effectiveness of treatments after they hit the market. The startup more than tripled its funding total in March following a $220 million fundraise that it says will be used for bulking up its software features even further.
.
As researchers working on Google Translate, Spence Green and John DeNero were initially surprised to learn that the search giants localization teamtasked with branding products for different areas of the worlddidnt use the tool. Anyone whos tried the machine translation program knows the results can be literal and awkwardinsufficient for many business use cases. We started building human-machine systems that utilize the scalability and efficiency of machines and the ingenuity and creativity of people, says Green. He describes Lilt as the world's first and only interactive, self-learning neural machine translation system. The companys technology-enabled translation services are used by enterprises and government entities around the world, including Intel and the U.S. Air Force.
.
Financial reporting can be laden with errors and omissions, whether intentional or not. MindBridges analytics software functions as an audit machine that searches financial documentation for missing links, incorrect data and even fraud. Its AI is trained on reliable accounting practices in order to identify unusual transactions and outliers. Founder Solon Angel, who left San Francisco for Ottawa after the 2008 financial crisis, partially credits the Canadian government for helping MindBrdige take shape. Now, established institutions like the Bank of England and the National Bank of Canada, as well as more than 8,000 firms and over 120 universities employ MindBridges software. In April, the startup was awarded a patent for data ingestion by the U.S. Patent and Trademark Office.
.
MindsDB is an open-source automated machine learning platform made for data scientists and developers to quickly train and deploy models. Our mission is to put machine learning in the hands of more people, says chief operating officer Adam Carrigan. Through MindsDB, data can be used straight from the database, datastores or business intelligence tools to generate AI-driven forecasts for what matters most to the business. Through its simple interface, users can train and deploy machine learning models directly in the database with a few code lines, standard query language or a few clicks. The open-source tool has more than 10,000 users; in late 2020, it began offering paid premium services.
.
Moveworks launched with a bot that could autonomously resolve employees IT issues. Its natural language processing capabilities allow it to understand conversational or ambiguous questions. A query such as I can't log in! will prompt Moveworks to automatically reset the password or multi-factor tokens. I spilled coffee on my laptop and now it wont turn on returns a filled-out loaner laptop form. CEO Bhavin Shah says the company celebrated a major landmark in the summer of 2018 when one customer resolved 20% of IT issues autonomously, without any human support. That number is now up to 40% on average, he says, and as high as 65% for one company. In March, Moveworks expanded its AI capabilities beyond IT to human resources, facilities and finance.
.
Narrativ was started with the idea that brands can add more customers by paying content creators who reference them in order to drive readers to their products. The companys platform also matches creators with the most relevant products and brands, from customers including Best Buy, Ulta Beauty, Samsung, LOreal, Yeti and Sephora. Despite the challenges of Covid-19, Narrativ doubled in size in 2020, and did so while prioritizing diversity. Its workforce is 60% people of color and 42% female.
.
Nines helps physicians and radiologists diagnose diseases faster. For example, its AI-based tools can measure lung nodules that accelerate diagnoses of certain respiratory diseases. This reduces the amount of time radiologists spend measuring pulmonary nodules and speeds the diagnosis for patients. By saving precious time otherwise spent on administrative and non-diagnostic tasks, the company says its technology allows imaging centers and hospitals to turn patients around faster.
.
Despite the oft-heard customer service disclaimer that calls may be recorded for quality and training'' that rarely happens, says chief revenue officer Sharath Keshava Narayana. Contact centers only listen to 1% to 2% of calls or even fewer. The startups founders saw a rich opportunity to apply speech analytics to this untapped opportunity, leveraging speech-to-text and natural language processing to find points of interest in human conversation. Observe.AI provides call center customers with supervised machine learning to better understand customer moods, if contact center guidelines are being followed and help coach human agents to provide a better customer experience.
.
Replicant is building artificial intelligence for customer service calls. The Atomic Labs alumnus, however, differs in its approach to call center automation. True to its namean homage to the genetically engineered humans in Blade RunnerReplicants solution involves an AI bot with a humanlike voice that can hold a conversation with people calling with customer service questions. Replicants voice AI agent eliminates wait times and can autonomously resolve basic issues, while routing calls that require high empathy to human reps. Launching its product just months before the pandemic, Replicants product was put to the test immediately, says CEO Gadi Shamia, and in one case scaled more than 30,000 AI-powered calls per day within 10 weeks. Every time I wait on a long hold, listening to the hold message on a loop, I am reminded why we started Replicant and how much work is still ahead of us, he says.
.
Digital training platform Sama was created in 2008 by Leila Janah, who wanted to connect students and people in developing countries to the digital economy and tech-oriented jobs. She was inspired to do so when she was just 25, after a stint teaching English in Africa. Since then, Sama has expanded significantly, developing training programs for corporate giants including Walmart, Google and NVIDIA. Its training data powers machine learning algorithms for an array of applications, spanning robot-assisted surgery to autonomous vehicles to personalized online shopping. Sama also launched an AI bias detection solution and, though Janah died of cancer in 2020, remains committed to improving job opportunities for people from disadvantaged communities, according to CEO Wendy Gonzalez.
.
Samsara was created with the idea that cloud and AI tools can make industrial operations safer, more efficient and more sustainable. The companys platform, which has proven to be particularly helpful for trucking fleet operators, collects information from real-time HD video feeds, sensors and data entry workflows and then uses machine learning to sift through the data to identify areas for improvement. For example, its dash cameras spot distracted driving and provide driver alerts and coaching tips in real time. Its platform monitors speeding and fuel consumption and is used to help cities like Boston manage electric vehicle fleets.
.
Alex Wang dropped out of MIT to launch Scale AI in 2016. Today he leads a unicorn that helps companies like Etsy, PayPal, Samsung and Toyota, as well as the Department of Defense, build and manage their AI and machine learning models. Scales focus is better labeling of vast amounts of data companies collect and need to train ML algorithms. In April, Scale closed a $325 million Series E investment round to expand its team and product offerings and added Jeff Wilke, former CEO of Amazon Worldwide Consumer, as an advisor.
.
Follow this link:
AI 50 2021: Americas Most Promising Artificial ...