Clarifai Inc. is an artificial intelligence (AI) company that specializes in computer vision and uses machine learning and deep neural networks to identify and analyze images and videos. The company provides a platform for building and deploying AI models, with a focus on visual recognition tasks such as object detection, content moderation, and visual search. Clarifai serves a range of industries, including media, retail, and government, and offers both cloud-based and on-premise deployment options.
Clarifai is headquartered in Wilmington, Delaware, with satellite offices in San Francisco, Washington, D.C., New York City, Tallinn, Estonia, Canada, and India. The company's technology is built on Deep learning and Neural network architectures, particularly Convolutional neural networks, which enable computers to learn from data examples and draw their own conclusions for predicting tags on images or videos.
History
Clarifai was founded in 2013 by Matthew Zeiler, a Ph.D. student who had recently placed in the top 5 spots of the 2013 ImageNet Challenge, a major competition in the field of Computer vision. Initially, the company offered free and paid versions of image and video recognition via an API, along with a consumer-facing iPhone app called Forevery. In 2014, Style Me Pretty, a wedding lifestyle website, became Clarifai's first customer, using the technology to provide images personally adapted to users.
In 2016, Clarifai released version 2 of its API, adding custom training and visual search capabilities. The following year, the company moved all research work to a San Francisco office and all government-related endeavors to a Washington, D.C. office. Later in 2017, Clarifai announced a mobile SDK that allowed users to run the platform without an internet connection. In 2018, the company released an on-premise solution, and in 2019, it opened a new office in Tallinn, Estonia.
Funding
Clarifai's funding history reflects its growth from a startup to an established AI company. In 2015, the company raised $10 million in its Series A funding round, led by Union Square Ventures. After the 2016 launch of its v2 API, Menlo Ventures led a $30 million Series B round, with participation from Union Square Ventures, Lux Capital, and Osage University Partners. In October 2021, Clarifai closed a $60 million Series C funding round led by New Enterprise Associates (NEA), with participation from existing investors Menlo Ventures, Union Square Ventures, Lux Capital, LDV Capital, Corazon Capital, and NYU Innovation Venture Fund, as well as new investors CPP Investment Board, NextEquity Partners, SineWave Ventures, and Trousdale Capital. This brought the company's total funding to $100 million.
Technology
The core of Clarifai's technology is based on convolutional neural networks, a type of Deep learning architecture that Zeiler focused on during his doctoral work. The platform enables computers to learn from data examples and make predictions, giving applications the ability to tag images or videos with relevant concepts. Clarifai's technology supports a variety of functions, including content moderation, visual search, visual similarity, and organizing media collections.
The platform includes pre-built recognition models for specific concepts like food, travel, and not-safe-for-work (NSFW) content, as well as a general model that can identify a wide range of concepts including objects, ideas, and emotions. Users can also create custom models to recognize specific objects, such as cars or breeds of dogs. In 2018, Clarifai released Model 1.5, which uses machine-labeled datasets and claims to recognize up to 11,000 concepts from object detection, as well as attributes like mood or theme.
Clients and Adoption
By 2017, Clarifai had approximately 100 customers. The company's technology was used by notable organizations including Unilever, Ubisoft, and BuzzFeed, as well as companies producing medical devices and drones. These clients leveraged Clarifai's platform for tasks such as image tagging, content moderation, and visual search, demonstrating the breadth of applications for its computer vision capabilities.
Military Work and Ethical Concerns
In 2018, Zeiler disclosed that Clarifai was a participant in Project Maven, a U.S. Department of Defense program focused on using AI for military applications. The disclosure came in a blog post after a Wired story reported that a former employee had filed a wrongful termination suit. The suit alleged that she was dismissed for requesting that Clarifai disclose a 2017 server compromise to the Pentagon and other customers. Zeiler asserted in a company blog post on June 13, 2018, that the breach involved only an isolated research server and that customers were notified following an external security audit.
In late January 2019, several Clarifai employees posted an open letter on a company message board expressing concerns about the ethical implications of the company's work with Project Maven and its potential involvement with unethical uses of AI. This reflects broader societal concerns about the ethical use of Machine learning and Artificial intelligence technologies, which have grown alongside the pace of AI advancement.
Technology
The core of Clarifai's technology is based on convolutional neural networks, an architecture that Zeiler focused on during his PhD work. The platform includes capabilities for content moderation, visual search, visual similarity, and organizing media collections. It provides pre-built recognition models that can identify specific concepts such as food, travel, or not-safe-for-work (NSFW) content. The general model identifies a range of concepts including objects, ideas, and emotion. Custom models allow users to recognize arbitrary objects, such as specific car models or dog breeds. In 2018, Clarifai released Model 1.5, which claims to recognize up to 11,000 concepts from object detection, as well as attributes like mood or theme, using machine-labeled datasets.
Clients
By 2017, Clarifai had about 100 customers. Its technology was used by companies such as Unilever, Ubisoft, and BuzzFeed, as well as by producers of medical devices and drones. The platform's applications included content moderation, visual search, and media organization, serving both commercial and enterprise needs.
Military work
In 2018, founder Matthew Zeiler disclosed that Clarifai was a participant in Project Maven, a united-states-department-of-defense AI program. The disclosure came in a blog post after a Wired story reported that a former employee had filed a wrongful termination suit, alleging she was dismissed for requesting that Clarifai disclose a 2017 server compromise to the Pentagon and other customers. Zeiler asserted in a company blog posted on June 13, 2018, that the breach involved only "an isolated research server," and that customers were notified following an external security audit. In late January 2019, several Clarifai employees posted an open letter on a company message board expressing concerns about the implications of Clarifai's involvement with potentially unethical uses of AI technology, such as automated warfare. This incident reflects broader societal concerns about the ethical development and use of Artificial intelligence.