Pecan AI nabs $66M to simplify AI development in the enterprise
Artificial intelligence startup Pecan AI Ltd. today announced that it has closed a $66 million funding round led by Insight Partners.
S-Capital, GGV, Dell Technologies Inc.’s venture capital arm, Mindset Ventures and Vintage Investment Partners participated in the Series C round as well. Pecan AI’s total outside funding now stands at $117.5 million.
Israel-based Pecan AI provides a platform that enables companies to build AI models for tasks such as predicting customer demand and finding opportunities to reduce operational costs. Historically, building a custom neural network required specialized technical know-how, as well as a considerable amount of time and effort. Pecan AI says that its platform allows companies to deploy AI models in weeks without the need for specialized machine learning expertise.
The most time-consuming aspect of machine learning projects is often not the AI development process, but rather the task of assembling a training dataset. A training dataset is a collection of records similar to the information that a neural network will be expected to process in production. The records are used to increase the neural network’s accuracy before it’s deployed.
A company building an AI model to predict store sales, for example, might use a collection of historical store sales logs as its training dataset. Because business data is often scattered across multiple business units, assembling a training dataset may require collecting information from several different systems. Pecan AI’s platform simplifies the task with connectors that can automatically retrieve information from a company’s systems of record.
The next step in machine learning projects is turning the collected training data into a form that the AI model can process. Pecan AI’s platform promises to help with that task as well. The startup provides a drag-and-drop interface for performing data preparation tasks. The platform also promises to automate feature engineering, or the process of determining what factors an AI should consider when making a decision.
When an AI analyzes a dataset to find a certain insight, such as how much revenue a store can be expected to generate in the next quarter, it more often than not doesn’t analyze all the information in the dataset. Instead, neural networks prioritize the most relevant factors. An AI tasked with forecasting store sales might take into account how much revenue a retail location generated in the previous 30 days, but not what products customers purchased.
Engineers usually have to specify manually what factors an AI should consider when making a decision. The task historically required specialized machine learning expertise. Pecan AI’s platform, in contrast, allows users to perform feature engineering even if they don’t have extensive familiarity with machine learning technologies.
Once a machine learning project’s training dataset is ready, Pecan AI automatically deploys several pre-packaged neural networks that could potentially be suitable for the customer’s use case. The startup’s platform runs a competition in which the neural networks are compared with one another. After the competition is complete, Pecan AI identifies the AI that proves most suitable for the project.
Enabling business analysts without extensive AI expertise to create customized neural networks has several potential benefits for companies. One is that firms without an in-house machine learning team gain the ability to deploy more sophisticated AI software than they otherwise could.
A second benefit, which is also useful for companies that do have dedicated machine learning staff, is that automating AI engineering tasks requiring specialized know-how speeds up the development process. The result, in theory, is a faster return on investment. Pecan AI says that one e-commerce company achieved a revenue increase within 14 days of deploying a neural network created with its platform.
“We believe that any company should be able to deploy AI-based predictive analytics, even without data science resources on staff,” said Pecan AI co-founder and Chief Executive Officer Zohar Bronfman. “This new funding will help us scale Pecan further to overcome the data science scarcity gap, enabling our customers to move beyond outdated data-mining techniques that offer little value in predicting future outcomes.”
Pecan AI says that its platform is used by several dozen companies, including Johnson & Johnson and other large enterprises. The startup told VentureBeat that its revenues have tripled over the past year but didn’t share absolute numbers. As for the coming year, Pecan AI has plans to double its headcount.
Image: Pecan AI
A message from John Furrier, co-founder of SiliconANGLE:
Your vote of support is important to us and it helps us keep the content FREE.
One click below supports our mission to provide free, deep, and relevant content.
Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.