MetaMind Launches State-of-the-Art AI Platform Bringing Automated Predictions and Smarter Decision-Making to Businesses

Raises $8 million in initial funding from Khosla Ventures and Marc Benioff to implement easy-to-use deep learning applications for Fortune 500 companies

PALO ALTO, Calif., Dec. 5, 2014 — (PRNewswire) —  Today, MetaMind launched its state-of-the-art artificial intelligence (AI) platform and announced initial funding of $8 million from Khosla Ventures and Salesforce chairman and CEO, Marc Benioff.  The company's ground-breaking deep learning software combines natural language processing and computer vision to solve a broad range of complex data problems, enabling automated predictions and smarter decision-making for businesses with greater speed and accuracy than ever before.


The company's breakthrough technology, based on co-founder and CTO Richard Socher's deep learning research, already outperforms many other systems on several public benchmarks including semantic relatedness, sentiment analysis and question answering. It has been widely recognized by peers and industry leaders as a major leap forward in improving accuracy of results in natural language processing. In addition, the MetaMind computer vision system outperforms all other methods on food classification and performs just as well as the winning entry of the 2014 ImageNet competition. 

"Textual and image data is growing exponentially and is already at a scale where human beings are no longer capable of processing it effectively. Recent and rapid advances in machine learning--deep learning in particular--are making solutions possible for problems that previously couldn't be addressed," said Sven Strohband, MetaMind CEO and CTO at Khosla Ventures. "We are pushing beyond established benchmarks in vision and language understanding and making this leading-edge technology broadly available to business users via our AI platform."

Currently, MetaMind works with both Fortune 500 companies, as well as smaller organizations across many industries to implement smarter, faster and more accurate decision-making solutions.

"MetaMind has developed some of the most impressive technology I have ever seen. Richard, Sven and the team are already delivering impressive results with real customers," said Marc Benioff. "The company's deep learning technology will have an enormous impact in multiple industries and has the potential to provide a generalized mathematical model for building machine intelligence that could be adapted for almost any discipline."

The company's general purpose AI platform can easily be applied to a variety of real-world problems. For instance, in health: classifying cancer in mammograms, finding anomalies in X-ray images or even presenting relevant health and nutrition information from identifying individual food images. In finance, MetaMind is able to accurately identify sentiment within financial reports to assist in algorithmic trading. Other applications include automating customer support responses on behalf of companies that see high volume of inquiries, automating social media analyses and even automatically calculating the fair price of a car from processing images of the vehicle. 

"I am very impressed by the depth of skills MetaMind has assembled, and its fast progress to match and even outperform the field on machine learning benchmarks to date," said Vinod Khosla, founder of Khosla Ventures. "The team's breakthrough work on deep learning algorithms and architecture is remarkable, and I'm excited by the ongoing brainstorming into the myriad of possible use cases that could make machine learning broadly accessible to high-value business users."

"MetaMind is one of the few deep learning startups with recognized and strong academic credentials in the deep learning research community, in both areas of visual data and natural language (and their combination), as well as regarding algorithms and architectures," said Yoshua Bengio,  one of most well known deep learning researchers. "They have achieved state-of-the-art performance on difficult academic benchmarks in both of these areas and are committed in advancing the research in difficult and exciting challenges for deep learning ahead of us."

"MetaMind has great people doing distinctive research at the intersection of deep learning and natural language processing," said Chris Manning, professor of computer science and linguistics at Stanford University. "They are also very experienced in building NLP solutions and can pragmatically combine this research with well-established approaches, as seen by their impressive benchmark results on sentiment analysis and semantic similarity. MetaMind can successfully bridge exciting new deep learning research in language and vision with easy-to-use applications."

MetaMind Smart Modules

MetaMind offers businesses a set of Smart Modules, which convert unstructured or semi-structured information such as pictures and sentences into mathematical objects from which they can make decisions with little or no assistance from humans. 

Examples of  Smart Modules are publicly available on the MetaMind Labs homepage, They include:

  • General Object Classifier: Classify images into thousands of visual categories
  • Food Classifier: Snap an image of food and the classifier identifies it
  • Image Classification Made Easy (IcMe): Train your own classifier for any problem and set of categories via "Drag, Drop & Learn" in the web browser
  • Textual Similarity: Identify and match similar strings of text based on semantic similarity
  • Sentiment Analysis: Identify positive, neutral or negative sentiment in text or social media (for example on Twitter which is integrated in our demos)
  • Text Classification (etcML): Train your own classifier on any problem via "Drag, Drop & Learn" in the web browser

Deep Learning Research at MetaMind

MetaMind continues to push research and performance on a variety of different benchmarks and explores applications of deep learning to new tasks and domains. The team has already achieved  best in class performance in four out of five benchmarks:

Natural Language Processing:

  • Semantic Relatedness
    2014 SemEval Competition Task 1: MetaMind obtains 83.4, a higher correlation with humans than the winning entry ( 82.8)
  • Paraphrase Detection
    Microsoft Research Paraphrase Corpus: Socher et al. 2011
  • Sentiment Analysis
    Stanford Sentiment Treebank: Socher et al. 2013 (EMNLP)

Computer Vision:

About MetaMind

MetaMind provides simple solutions for complex business problems by leveraging AI to tap into large volumes of data to make automated predictions and smarter decisions with more speed and accuracy than ever before. The company's advisors include Vinod Khosla, founder of Khosla Ventures; Marc Benioff, CEO of Salesforce; Professor Chris Manning, Stanford department of linguistics  and computer science; and Professor Yoshua Bengio, Université de Montréal department of computer science and operations research. Founded in 2014, MetaMind is based in Palo Alto, Calif. and currently has 10 employees. For more information, please visit:

Logo -

To view the original version on PR Newswire, visit:


For Media Inquiries: Jacqueline Tanzella
Email Contact 415-741-7848

Review Article Be the first to review this article
Latest Blog Posts
    Administrator registration is locked (again)
    May 20th, 2022 by Administrator
    Even though we found a way to filter a lot of the fake registration, there are too many fake account created yesterday. Next week we will going to change to more strict restriction, and add new one. Of course, it will make the registration …
    Susan Smith HP Announces HP Z240 Tower and Z240 SFF Workstations
    October 2nd, 2015 by Susan Smith
    Jeff Wood - vice president, WW Product Management,  Workstation and Thin Client Business HP and Josh Peterson- director, WW Product Management,  Workstation and Thin Client Business , HP met with the press in a virtual press briefing prior to the …

© 2022 Internet Business Systems, Inc.
670 Aberdeen Way, Milpitas, CA 95035
+1 (408) 882-6554 — Contact Us
ShareCG™ is a trademark of Internet Business Systems, Inc.

Report a Bug Report Abuse Make a Suggestion About Privacy Policy Contact Us User Agreement Advertise