10 Cool Machine Learning Startups To Watch
Machine learning is technology that trains software so developers don't have to code it by hand. The number of new companies in the category has grown exponentially over the past few years. Here are 10 machine learning startups worth a closer look.
Machine learning companies are being snapped up in droves by tech giants cognizant that these startups represent a new wave of technology innovation.
This month alone, Intel announced plans to acquire deep learning startup Nervana Systems. Apple confirmed it would acquire Turi. Earlier this year, Twitter acquired Magic Pony Technology, Salesforce acquired PredictionIO, ESI Group acquired Mineset, and Apple acquired Emotient, among other deals.
PricewaterhouseCoopers said 29 machine learning companies have been acquired so far this year by companies large and small, and total deals in 2016 will likely exceed the 37 such buyouts made last year.
These deals represent a fraction of the action happening in machine learning today. Big companies like Google, Microsoft, Facebook, eBay, LinkedIn, and plenty of others, already have advanced internal projects underway.
[How are companies using machine learning today? Read 11 Cool Ways to Use Machine Learning.]
There are hundreds of startups in this area. Although machine learning's concepts have been around for decades, a handful of factors are driving its rise to prominence right now. These include the huge volumes of data generated by the internet, the internet of things (IoT), social media, sensors, and more. Another driver is the continued decline in storage costs, which makes it economically feasible to keep all that data.
"Ten years ago, we struggled to find 10 machine learning-based business applications. Now we struggle to find 10 that don’t use it," Alexander Linden, research vice president at Gartner, said recently.
Gartner defines machine learning as "a technical discipline that provides computers with the ability to learn from data (observations) without being explicitly programmed."
The massive amount of data available today lets software be "trained," rather than programmed.
Machine learning is being used across industries including finance, retail, marketing, healthcare, cyber-security, and even agriculture. There are companies specializing in machine learning applications for these verticals.
There are machine learning startups casting a wider net and appealing to an audience of data scientists and developers who then turn around and create technology for these vertical industries or for their own enterprises.
With so many interesting and innovative machine learning startups sprouting up every day, it's hard to know which ones are worth a closer look. We're aiming to offer an interesting -- though by no means comprehensive -- look at 10 newcomers.
If you know of others doing innovative things or working with unusual industries, please share them in the comments section below.
Lukas Biewald, a former lead data scientist at Yahoo, founded CrowdFlower to build training data for machine learning. Training data is used by data scientists to teach their algorithms to learn.
The company website says it is focused on making data useful by helping teams collect, clean, and label data at scale. Crowdflower raised a new $10 million round of funding in June 2016 that included Microsoft as an investor. The funding will be used to fuel adoption of CrowdFlower AI, launched last year, which enables machine learning algorithms to go beyond prediction to provide judgment on how likely the prediction is to be correct -- known as a confidence level.
H2O.ai, cofounded by CEO SriSatish Ambati, offers an open source machine learning platform. It's designed to enable organizations to run a series of algorithms on all their stored data, whether it's stored in Hadoop, Spark, Excel, or something else.
Darktrace is a three-year-old UK-based machine learning company focused on cyber-security. The company is using machine learning technologies to monitor network traffic and events to look for anomalies that may indicate attacks. IT is alerted to suspicious activity, and if staff cannot respond immediately, the software itself can respond to a potential threat. The company recently closed a new $65 million round of funding.
LogDNA is an early stage company aiming to apply machine learning to IT log data for a predictive approach to managing and maintaining systems. The idea is to help organizations detect and address IT problems before they happen. As IT systems have grown more complex, organizations need to not only look at infrastructure logs, but also custom application event logs, database logs, mail server logs, operating system logs, and more. LogDNA's agent is designed to allow organizations to specify the logs to watch, and lets administrators set up views and alerts based on custom smart filters. The company raised $1.6 million in July 2016.
Amplero's predictive customer lifetime value platform, AmpIP, is designed to give marketers the power of machine learning. For instance, organizations can set boundary rules and constraints and then unleash the platform to test thousands of permutations to ensure the right message is delivered to the right person at the right time. Amplero raised an $8 million round of venture capital in July 2016 with investors including Salesforce Ventures.
Farmers Business Network describes itself as an independent farmer-to-farmer network built by and for farmers. Member farmers receive analytics on a variety of metrics such as seed performance, agronomic practices, input prices, and yield benchmarking. The analyses are based on more than 55 million acre-events of real-world data from the FBN community of farms. The network also uses machine learning to improve data quality, discarding data points and results that appear abnormal based on results from similar geographies and growing conditions. Google Ventures is among the investors backing this organization.
DataRobot automates the data science process. The system lets you upload your data and tell it what you want to predict. It automatically and instantly creates hundreds of models. The platform integrates open source machine learning algorithms from R, Python, Spark, H2O, and others, according to the DataRobot website. Earlier this year, the company raised $33 million in a round led by New Enterprises Associates that included Intel Capital, Accomplice, and IA Ventures.
Sentenai is a cluster-native stream database that uses machine learning to predict structure and query patterns in data streams. Primary applications are in IoT and logistics, according to the company's website. Based in Boston, Sentenai is at an early stage. It raised $1.8 million in seed funding in January.
FiveAI is another UK-based startup focused on machine learning and AI. But this one is using the technology to create the intelligence for tomorrow's autonomous cars. The company raised $2.7 million in July 2016. FiveAI's website says it is using AI to enable autonomous vehicles to drive safely in complex urban environments without the need for driver involvement. The company says its approach is a departure from the ways such systems have been created to date, and uses highly accurate and dense 3D maps.
FiveAI is another UK-based startup focused on machine learning and AI. But this one is using the technology to create the intelligence for tomorrow's autonomous cars. The company raised $2.7 million in July 2016. FiveAI's website says it is using AI to enable autonomous vehicles to drive safely in complex urban environments without the need for driver involvement. The company says its approach is a departure from the ways such systems have been created to date, and uses highly accurate and dense 3D maps.
-
Read more about:
Technology StartupsAbout the Author(s)
You May Also Like