Whether you’re a tech enthusiast or not, you’ve probably heard about the latest AI news. And there are plenty of great stories out there. But there are also a few not-so-great stories. Here’s what you need to know about these AI startups.
Using advanced AI technology, Lily AI is a ‘product attributes’ platform that enables retailers to configure ten times more attributes for each product. These enriched attributes help retailers to supercharge their demand forecasting and site search conversion, as well as improve search engine optimization and merchandise planning.
Lily’s technology uses image recognition to extract more than 15,000 product attributes, and these data points are then enriched across the retail stack. For example, the AI-powered system can help retailers identify consumer preferences for specific attributes of a product. The resulting “product-taxonomy” is used to train the algorithms that perform product search. The company plans to expand its solution to include additional applications within the retail stack.
The startup has already signed big name customers, such as The Gap, Macy’s, Bloomingdale’s and thredUP. The company plans to expand its business into other retail verticals, such as home and beauty.
Founded in 2014, Wallaroo is an enterprise AI platform designed to help companies deploy machine learning models in production environments. The company aims to make the last mile of the ML process easy, secure and scalable. Wallaroo will use the funding to further the development and sales of its platform.
Wallaroo’s platform consists of four key components: an MLops engine that runs on-premises or in the cloud, a distributed processing engine, an audit and performance metrics tool and data connectors. Each of these components is designed to help data scientists achieve visibility into live models. The platform also provides observability features that improve model validation checks. In addition, Wallaroo has tools to address common data drift issues. This helps users identify model changes faster.
Wallaroo’s platform can run multiple ML models on shared infrastructure. This allows users to reduce computing costs by 80%. Wallaroo’s observability tools provide compute and model performance metrics, including audit logs and A/B testing comparisons. Moreover, Wallaroo’s data connectors allow for custom integrations with in-house solutions and enterprise data sources.
Mostly AI, a startup in Austria that generates artificially generated synthetic data, just raised a $25 million Series B round led by Molten Ventures. The company’s synthetic data technology reduces the time it takes to generate data by 90 percent. Mostly AI’s products are aimed at companies involved in software testing, quality assurance and data management.
The company has already secured customers from several Fortune 100 companies, as well as multiple insurers. Its flagship product, the MOSTLY AI 2.0, is a software platform that automatically synthesizes complex data structures. It also adapts to the data structures of companies using it. The company plans to hire fresh talent and expand its presence in North America and Europe. Its chief aim is to make its products as accessible as possible to software engineers, quality testers and data scientists.
Using artificial intelligence and deep learning, Verusen’s AI-powered supply chain management platform offers the simplest way to manage materials. It also helps to increase efficiencies while reducing costs. The company is now serving 25 countries around the world, with the goal of reaching 100 employees by the end of the year.
Verusen’s $25 million Series B funding round, led by Scale Venture Partners, will enable the company to build out its AI-powered technology platform. The company has been able to achieve 10x growth over the past year and has also landed major partnerships with SAP, Accenture and NTT Data GSL. The company has also opened a new headquarters in Atlanta’s Tech Square.
Verusen has also made a name for itself in the AI-powered supply chain management space. The company offers a supply chain management platform to help enterprise clients manage risk, optimize inventory, and reduce costs.
Among the most hot topics within the AI Discoveryhatmakertechcrunch market is Machine Learning Operations, which aims to streamline the continuous development of models at scale. This space has been expanding in recent years, and the market is expected to grow to $4 billion by 2025.
The machine learning stack is getting faster and faster, and new startups are addressing a wide range of machine learning problems. Some are working on generic datasets, while others are developing AI products for particular niches.
MLOps is a set of practices and tools that allow organizations to deliver trusted decisions in real-time. In a nutshell, MLOps involves standardizing the collaboration between data scientists and operations. This opens the door to better governance and faster model operationalization.
Another important part of the new MLOps workflow is data and modeling versioning. This allows teams to label data and create structured embeddings, which can be fed into Watson Studio ML algorithms.