Home >Technology peripherals >AI >Year-end review: The ten hottest data science and machine learning startups in 2022
As businesses deal with ever-increasing amounts of data, both generated internally within the organization and collected from external sources, finding effective ways to analyze and “manipulate” this data for competitive advantage becomes increasingly important. The more challenging it is.
This has also driven the need for new tools and technologies in the field of data science and machine learning. According to the "Fortune Business Insights" report, the global machine learning market will reach US$15.44 billion in 2021 and will reach US$21.17 billion this year. It is expected to grow to US$209.91 billion by 2029, with a compound annual growth rate of 38.8%. .
At the same time, according to a report by Allied Market Research, the global market size of data science platforms in 2020 was US$4.7 billion and is expected to reach US$79.7 billion by 2030, with a compound annual growth rate of 33.6%.
“Data science” and “machine learning” can sometimes feel a little confusing or even used interchangeably. These are actually two different concepts, but they are related because data science practices are key to machine learning projects.
According to the definition of the Master's in Data Science website, data science is a field of study that uses scientific methods to extract meaning and insights from data, including developing data analysis strategies, preparing data for analysis, developing data visualizations, and building data models .
According to the "Fortune Business Insights" report, machine learning is a subsection of the broader field of artificial intelligence and refers to the use of data analysis to teach computers how to learn (that is, to imitate the way humans learn) Use algorithms and data-based models.
The demand for data science and machine learning tools has spawned a number of startups developing cutting-edge technologies in the field of data science or machine learning. Let’s take a look at 10 of them:
Aporia develops a full-stack, highly customizable machine learning observable A comprehensive platform that data science and machine learning can use to monitor, debug, interpret, and improve machine learning models and data.
Aporia was founded in 2020 and received an additional $25 million in Series A funding in March 2022, 10 months after receiving $5 million in seed funding.
Aporia will use the funding to triple the size of its workforce by early 2023, while also expanding its U.S. presence and expanding the range of use cases its technology covers.
Baseten was officially released in April this year. It provides products that can speed up the process from machine learning model development to production-level applications.
According to Baseten, the technology, which has been in internal beta since the summer of 2021, can automate many of the skills required to bring machine learning models into production, helping data science and machine learning teams integrate machine learning into In the business process, no back-end, front-end or MLOps knowledge is required.
Baseten was founded in 2019 by CEO Tuhin Srivastava, CTO Amir Haghighat and chief scientist Philip Howes, who all previously worked at e-commerce platform developer Gumroad. Baseten raised $12 million in Series A funding in April this year, following an earlier $8 million in seed funding.
Deci has developed a deep learning development platform for building the next generation of artificial intelligence and deep learning applications. Deci’s technology is designed to help address the “AI efficiency gap,” where computer hardware is unable to keep up with the demands of machine learning models that are growing in size and complexity.
Deci’s platform takes production into account early in the development lifecycle, helping data scientists bridge this gap and reduce the time and cost of resolving issues when deploying models in production. According to Deci, the platform incorporates Deci’s proprietary AutoNAC (Automatic Neural Architecture Construction) technology to provide a “more efficient development paradigm” to help AI developers leverage hardware-aware “neural architecture search” to build deep learning models. to meet specific production demand targets.
Deci was founded in 2019 and raised US$25 million in Series B financing led by Insight Partners in July this year. Just 7 months ago, Deci had just received US$2,100 in Series A financing. Ten thousand U.S. dollars.
Galileo has developed a machine learning data intelligence platform for unstructured data, allowing data scientists to inspect, discover and fix critical machine learning errors throughout the machine learning lifecycle.
In early November this year, Galileo launched the Galileo Community Edition, a free version of the platform, allowing data scientists engaged in natural language processing to build models more quickly using higher-quality training data.
Galileo came out of stealth mode when it received US$5.1 million in seed funding in May this year, and then on November 1, it received US$18 million in Series A financing led by Battery Ventures. Galileo's co-founders include CEO Vikram Chatterji, who was the head of cloud AI project management at Google; Atindriyo Sanyal, a former software engineer at Apple and Uber; and Yash Sheth, a former software engineer on Google's speech recognition system engineer.
Founded in 2021, Neuton develops an automated no-code “tinyML” platform and other tools for developing tiny machine learning models that can be embedded in microcontrollers to enable the edge Devices become smart.
Neuton's technology is being used in a wide range of applications, including predictive maintenance of compressor water pumps, preventing grid overload, room occupancy detection, handwriting recognition on handheld devices, transmission failure prediction and water pollution monitoring equipment.
The vector database and search technology developed by Pinecone mainly provide support for artificial intelligence and machine learning applications. In October 2021, Pinecone launched Pinecone 2.0, taking the software from research laboratories to production applications.
Pinecone was founded in 2019 and was officially launched last year. It received US$10 million in seed round financing in January 2021 and US$28 million in Series A financing in March this year.
In October this year, Pinecone expanded its machine learning search infrastructure portfolio with the launch of a new “vector search” solution that combines semantic and keyword search capabilities.
Gartner named Pinecone a "Cool Vendor" in the field of artificial intelligence and machine learning data in 2021.
In May, Predibase came out of stealth mode with a low-code machine learning platform that allows data scientists and non-experts to quickly develop applications with “best-in-class” machine learning infrastructure. Machine learning model. The software is currently in beta use at many Fortune 500 companies.
Predibase offers its technology as an alternative to traditional AutoML to develop machine learning models to solve real-world problems. The platform uses declarative machine learning, which Predibase says allows users to specify machine learning models as "configurations," or simple files, telling the system what the user wants and letting the system figure out the best way to meet that need.
Predibase was co-founded by CEO Piero Molino, Chief Technology Officer Travis Addair, Chief Product Officer Devvret Rishi, and Stanford University Associate Professor Chris Re. Both Molino and Addair once worked at Uber. While at Uber, the two developed the Ludwig open source framework for deep learning models, and the Horovod open source framework for extending and distributing deep learning model training to massive data (Predibase is built on Ludwig and Horovod.)
In May of this year, Predibase received US$16.5 million in seed and series A financing led by Greylock.
Snorkel was founded in 2019 and originated from the Stanford University Artificial Intelligence Laboratory, where the company’s five founders worked on solving the problem of lack of labeled training data for machine learning development Methods.
Snorkel developed Snorkel Flow, a data-centric system that became fully commercially available in March this year, by using programmatic tagging, a key step in data preparation and machine learning model development and training. ) to accelerate the development of artificial intelligence and machine learning.
Snorkel's valuation reached $1 billion in August 2021, when the startup secured $85 million in Series C funding, using the funds to grow its engineering and sales teams and accelerate its platform development.
Vectice develops an automated data science knowledge capture and sharing solution. Vectice's technology automatically captures the assets data science teams create for projects, including datasets, code, models, notebooks, runs, and illustrations, and generates documentation throughout the project lifecycle, from business requirements to production deployment.
Vectice’s software is said to be designed to help enterprises manage transparency, governance and alignment with AI and machine learning projects and deliver consistent project results.
Vectice, founded in 2020 by CEO Cyril Brignone and CTO Gregory Haardt, raised $12.6 million in Series A funding in January this year, bringing total funding to $15.6 million.
Verta develops artificial intelligence/machine learning model management and operation software that data science and machine learning teams use to deploy, deploy, and machine learning models throughout the entire artificial intelligence and machine learning model lifecycle. Operations, management, and monitoring inherently complex models.
In August, Verta enhanced the enterprise capabilities of its MLOps platform, including adding a native integration ecosystem, additional capabilities around enterprise security, privacy and access control, and model risk management.
Verta was founded in 2018 and officially released in 2020. This year, it was named a "Cool Vendor" in the core AI technology field by Gartner.
The above is the detailed content of Year-end review: The ten hottest data science and machine learning startups in 2022. For more information, please follow other related articles on the PHP Chinese website!