Data Explorer processes unlabeled visual data, boosting creation of production-ready AI models

AI-driven visual data platform Akridata has announced the launch of its flagship product Data Explorer in the Azure Marketplace. Data Explorer is designed specifically to process visual data in the machine learning (ML) life cycle, allowing data science teams to easily explore, search, analyze and compare visual data to improve datasets and model training. 

The Data Explorer platform offers virtual connections to multiple data sources, enables the exploration of visual data on unlabeled datasets, allows for image-based similarity searches, supports viewing model performance from multiple perspectives, and enables data comparison across numerous sets.

“One of our platform’s standout features is its ability to handle massive volumes of visual data without any performance issues or infrastructure limitations. This allows businesses to store and analyze data at scale without worrying about the usual headaches of managing large datasets,” Vijay Karamcheti, CEO and cofounder of Akridata, told VentureBeat. “With our secure and scalable platform, users can finally extract the insights they need to improve operations and gain a competitive advantage.”

By making Data Explorer available in the Microsoft Azure Marketplace, Akridata aims to provide a higher level of accessibility and ease of use for data scientists seeking insights from complex datasets and accelerate the path to building production-grade AI models.

“We are thrilled to have Data Explorer now available on the Microsoft Azure Marketplace,” said Sanjay Pichaiah, VP of products and GTM at Akridata. “With this partnership, we are amplifying global access to a cloud-based tool that helps data scientists explore, curate and use visual data at a large scale.”

“Azure offers an array of platform integrations, including Azure Data Factory, Azure Databricks, and Azure Synapse Analytics, that effortlessly integrate with Data Explorer,” Karamcheti said. “Customers would be able to derive even more value from their data by seamlessly incorporating our platform into their existing Azure-based data processing and analytics workflows.”

Akridata is also on the AWS marketplace. The company said being a standing AWS partner has allowed Akridata to reach a wider audience and expand its impact in the tech industry. 

Enhancing AI development pipelines

Data Explorer is designed to aid data science teams using visual data to improve datasets and model training. The company claims it is the first platform focused solely on processing visual data in the machine learning (ML) life cycle. 

“As volumes of visual data have exploded, the need to manage and select training sets has become paramount,” said Karamcheti. ”Data Explorer enables data scientists to quickly and easily explore, search, compare and analyze more than one million frames of visual data. By drastically reducing the time spent on data selection and curation, organizations can avoid wasting time on data labeling, and focus on accelerating their path to model accuracy.”

Karamcheti said another benefit of using the platform is its ability to explore visual data on unlabeled datasets by combining traditional metadata-based filtering with content feature–based latent-structure exploration. This allows users to better understand the dataset’s inherent clustering or segmentation structure.

The platform can also perform image-based similarity searches on millions of images in seconds, which can be further refined through interactive scoring on a subset of data to search for domain-specific features by combining active search techniques.

A data-centric way to manage visual data

Karamcheti believes that the key to managing the growth of visual data will be switching from a model-centric approach to a data-centric one. 

“Despite the ever-growing amount of visual data in our world, AI continues to rely on a model-centric approach. The problem with this is it was largely reliant on rules and heuristics. Data, rather, should be at the root of every decision made,” he explained. “The potential uses of visual data to improve real-world AI applications are huge only if we can find the algorithmic means to assess, store, curate and select visual data.”

The company said the platform addresses the issue of data privacy and security by providing users with granular control over access to data and compliance with regulatory requirements. It offers end-to-end data encryption in transit and at rest and integrates with existing authorization mechanisms to ensure secure access to data.

In addition, the company aims to be a leader in visual data analysis, offering seamless integration with existing workflows and tools, and providing customers with a comprehensive and powerful solution for managing and analyzing visual data.

“Advanced analytics capabilities such as computer vision and deep learning can help companies derive valuable insights from visual data,” said Karamcheti. “By unlocking the potential of visual data, we aim to empower businesses to make data-driven decisions confidently.”