These startups are applying artificial intelligence techniques to business intelligence, big data, cybersecurity, APM, autonomous vehicles, healthcare and more.
Headquarters: Ra'anana, Israel
Anadot's focus is on helping enterprises discover "business incidents" that could lead to costly financial losses or brand damage. It applies machine learning to traditional business intelligence, and the company says that its technology "notices - before you do and before your customers do - if something is off; if a metric is out of place." It adds, "The beauty of Anodot is that it identifies these anomalies at such an early stage that it can predict that something will become a major issue long before it does."
The analysts at Gartner have been impressed with Anodot, and they named it a Cool Vendor in Analytics for 2016. According to Crunchbase, the company has raised $27.5 million in funding.
Headquarters: Palo Alto, Calif.
While Anodot takes a broad approach, Simility takes a much more focused tack to business intelligence analysis. It helps financial firms address one of their most pressing problems: fraud. According to the company website, "Simility comes from the Latin 'similis,' meaning 'similar.' The power of our algorithms lies in separating like from unlike signals so that humans can create meaning from them."
The company was founded by a number of Google veterans, and Crunchbase says that it has raised $24.7 million to date.
Headquarters: Cambridge, UK
Darktrace has quickly become a powerhouse in the growing field of machine learning-enabled cybersecurity. According to the company, Darktrace technology now protects more than 4,000 networks and has found 53,000 previously unknown threats that might have gone unnoticed with other types of security solutions. Its technology was designed to mimic the immune system, and the company claims, "Our self-learning approach is the first non-consumer application of machine learning to work at scale, across all network types, from physical, virtualized, and cloud, through to IoT and industrial control systems."
Darktrace founders include mathematicians who previously worked at Britain's MI5 and GCHQ spy agencies. It has raised a whopping $179.5 million in funding.
Headquarters: MontrĂ©al, Canada
This machine learning startup offers a service that developers can use to "build smarter software." Fuzzy.ai creates smart agents that can understand "fuzzy logic" that allows developers to express ideas in regular language, like " People who buy a shirt in a particular style may like another shirt with the same style" or "Orders where the credit-card country are different from the delivery country are suspicious." The agent then applies those rules to your app and gets better over time. The technology is delivered as an API, with prices based on the number of API calls.
The Fuzzy.ai team is small - just five employees - but it has raised funding from 500 Startups, Real Ventures, Interaction Ventures and iNovia Capital.
Headquarters: San Francisco, Calif.
Like Fuzzy.ai, Instana is targeted at developers, more specifically at developers who are using a DevOps approach. This startup offers an application performance management (APM) solution that uses machine learning to improve application development and delivery in environments where the application is always changing due to agile development, continuous delivery and container orchestration. It supports a lot of the most popular programming languages and DevOps tools, including .NET, Java, Go, Node.js, PHP, Ruby, Scala, Ansible, Chef, DC/OS, Kubernetes, OpenShift and Puppet.
According to Crunchbase, Instana has raised $26 million in funding.
Headquarters: Bristol, UK
While most of the machine learning startups on this list offer software, Graphcore sells hardware (as well as some software). It makes intelligence processing units (IPUs), which are special processors designed to handle machine learning workloads very quickly. It also offers Poplar, a graph programming framework for use with its IPUs. The company claims that its IPUs and Poplar together can improve performance for AI workloads by 10 to 100 times.
Although this is one of the younger startups on our list, it is also one of the best funded, having raised $110 million in capital.
Headquarters: Tel Aviv, Israel
Panoply claims to offer "the world's only smart data warehouse." In a typical big data project, it might take dozens of engineers, developers and data scientists to get unstructured data from a data lake into a usable format and create the applications necessary to analyze that data. However, Panoply claims that its machine learning and natural language processing capabilities allow it to do most of that work on its own without any help from these experts. It handles schema building, data mining, complex modelling and performance tuning, freeing up valuable staff for more important tasks.
Crunchbase reports that the company has raised $14.3 million in funding so far.
Headquarters: Sunnyvale, Calif.
Crowdz brings together two of the hottest trends in technology: machine learning and blockchain. It says it is creating "the worldâ??s first blockchain-based B2B ecommerce marketplace." It uses artificial intelligence and machine learning to match up buyers and sellers so that companies can get access to the goods they need up to 10 times faster than through traditional methods. It is also using blockchain-based technology to create a smart transaction network.
Crowdz was founded by experienced supply chain and ecommerce veterans, including some who worked for Cisco. But so far, Crunchbase says it has raised only a little over $1 million in funding.
Headquarters: Mountain View, Calif.
As you might guess from the name, this startup is focused on machine learning for self-driving vehicles. Drive.ai's founders were researchers at the Stanford University Artificial Intelligence Lab. So far, it has raised $77 million in funding. It currently has a long list of job openings available and promises job seekers, "Besides the usual (competitive salaries, generous healthcare benefits), you can look forward to excellent perks which will include free lunches/dinners, social events, Q&A sessions over happy hour, and an open and collaborative environment. We offer significant equity packages which means we are all owners in the company, further driving our passion for success."
Headquarters: Paris, France
Many activities within the healthcare industry, particularly those related to diagnostics, are prime candidates for automation with machine learning technology. Cardiologs uses machine learning to read electrocardiogram (ECG) results. According to the company, traditional methods for reading ECGs are accurate only 59 percent of the time, but Cardiologs is accurate 91 percent of the time. The service isn't available to customers yet, but has been approved by the FDA.
According to Crunchbase, the company has raised $7.9 million in funding.
Artificial intelligence (AI) looks likely to be one of the most influential technology trends for 2018. And machine learning is poised to be one of the most aspects of AI that enterprises will need to master.
According to the latest forecast from IDC worldwide spending on cognitive and AI solutions, include machine learning solutions, is likely to achieve a compound annual growth rate (CAGR) of 50.1 percent from 2016 to 2021. In 2017, total revenue in the market was $12.0 billion, 59.1 percent higher than in 2016. By 2021, the analysts predict that revenues could hit $57.6 billion.
Large technology companies like IBM, Microsoft, Google, Apple, Facebook, Salesforce and Amazon have been investing heavily in AI and machine learning. They each have armies of researchers devoted to advancing machine learning capabilities, and they are also acquiring machine learning startups at a frantic pace.
For example, in 2017, Google bought machine learning competition platform Kaggle and India-based Halli Labs, which made machine learning to fix "old problems." Microsoft acquired Maluuba, which had impressive deep learning and reinforcement learning capabilities. Apple snapped up Lattice Data, which used machine learning to make unstructured data more structured, and Amazon reportedly purchased cybersecurity-focused machine learning startup Harvest.ai. And that's just a handful of the biggest machine learning acquisitions of the year.
With more software vendors and enterprises looking to add machine learning capabilities to their applications in 2018, the startup buying spree will almost certainly continue. And entrepreneurs continue to found new machine learning startups at a very rapid rate.
Which of these machine learning startups look the most promising? Here are ten that seem particularly noteworthy and worth watching in 2018.