Verta, the Operational AI company, today announced continued momentum in the first half of 2023, with strong adoption and continued enhancement of its Operational AI platform, the launch of a Free Tier program to expand access to Model Catalog for the Data Science and Machine Learning communities, and the release of its latest Verta Insights research study covering AI Regulatory Readiness.

Companies from Fortune 50 enterprises to startups have adopted Verta’s Operational AI platform and use Verta Model Catalog as their centralized control tower to manage machine learning assets throughout the ML lifecycle, from development, staging and production to archive. The Verta Operational AI Platform takes any ML model and instantaneously packages and delivers it using best-in-class DevOps support for CI/CD, operations and monitoring, while ensuring safe, reliable and scalable real-time AI deployments.

Recent enhancements to Verta Model Catalog include:

• Enterprise AI Ontology: Model Catalog enables organizations to define their enterprise AI ontology, incorporating customizable attributes, metadata, annotations and relationships for comprehensive knowledge representation. It serves as a standardized framework for running an AI-driven business, promoting knowledge sharing, aligning AI projects with objectives, ensuring governance and compliance, and facilitating seamless integration. The ontology supports effective communication, consistency and scalability in AI initiatives. It allows organizations to define and identify relevant models, for example, to identify all models in use that relate to generative AI or copyright materials, or that are subject to specific AI regulations like the EU AI Act or New York Law 144.
• Governance Workflow & Checklists: Easily customized checklists provide standardized controls that facilitate the smooth progression of models throughout the AI lifecycle, ensuring consistency, accountability and compliance while enabling effective management and evaluation of models. With Verta, organizations can assign checklists to roles and provide clear instruction to model stakeholders of what’s required to ensure completeness, documentation and adherence to business standards and IT policies. Serving as model acceptance criteria and logged for audit purposes, checklists promote traceability, accountability, coordination and collaboration, as well as facilitating the adoption of standards and enhanced transparency throughout the AI lifecycle.
• Lifecycle Activity Logging: Maintaining an activity log for an organization’s ML models is an essential part of model lifecycle management to support Responsible AI, regulatory compliance, and transparency and accountability in AI development. With Verta Model Catalog, organizations maintain comprehensive model activity logs, ensuring they have a documented history of their models’ development and enabling them to make informed decisions throughout the model lifecycle. Coupled with governance controls and an enterprise ontology, activity logging empowers organizations to establish a clear model chain of custody, demonstrate effective controls, ensure transparency and provide safety measures with guardrails for the responsible development and deployment of AI systems.

“Verta works closely with our customers and industry stakeholders to understand the key challenges that organizations face today in streamlining and optimizing their data science and machine learning operations. The latest enhancements in Model Catalog build on the core capabilities in the Verta platform for enabling operational excellence in AI/ML,” said Meeta Dash, Vice President, Product at Verta.

Free Tier Program Expands Access to Verta Model Catalog

Verta rolled out a Free Tier Program that enables data scientists to use, at no cost, Model Catalog to easily and securely manage all their machine learning (ML) models as the first step in their organizations’ journey toward Responsible AI and compliance with the growing number of AI regulations.

Verta Model Catalog lets data scientists publish all their models and model versions in a secure central catalog, along with all the relevant metadata, documentation and artifacts. After they register their models in the catalog, data scientists can share the models and related documentation with colleagues in other teams or departments in their organizations.

With one-click scanning of models for vulnerabilities, configurable governance checklists and centralized access to training data and detailed audit logs, Verta Model Catalog helps organizations ensure compliance with Responsible AI principles and comply with current and pending regulations like the EU AI Act and the various US federal and state requirements.

The Verta Free Tier Program makes it easy for data scientists to start organizing, managing and monitoring all their ML assets in a centralized hub. We also enable Data Science to collaborate more effectively with other stakeholders across the ML lifecycle, supporting operational excellence and enhanced AI governance,” said Manasi Vartak, Founder and CEO of Verta. “Free Tier is a commonsense on-ramp for organizations looking to streamline ML workflows, promote innovation and provide executive leadership with a 360-degree view of their ML model portfolio.”

Interested data scientists can sign up for the Free Tier Program at app.verta.ai.

Few Companies Prepared for AI Regulations, Verta Insights Study Shows

With generative AI like ChatGPT driving heightened regulatory activity around AI, a recent study by Verta Insights, the research practice within Verta, found that while a majority of organizations view AI regulatory compliance as a priority and believe that they will face an increasing number of AI regulations in the near future, few companies today are well prepared to meet current or future regulatory requirements.

Proposed laws like the EU AI Act and the US federal Algorithm Accountability Act, as well as current or proposed laws in US states like California, Illinois, New York and Texas, create new compliance and reporting requirements around companies’ use of AI and machine learning, intended to protect consumers against privacy violations, bias in automated decision-making and other potential harms.

Key insights from the AI Regulatory Readiness study included:

• AI regulations create a need for organizations to move from ad hoc approaches for documenting AI/ML systems, processes and models to an effective governance framework for overseeing the development, deployment and monitoring of AI models.
• Most companies currently recognize AI regulations as a priority, but to date only a small minority have put in place tools and teams to support compliance. That’s changing as laws like the EU AI Act make tools for regulatory compliance “must-haves.”
• Leaders view regulations as an opportunity to create competitive advantage by investing early in tools for ML monitoring, model governance and risk management, and model catalogs that drive efficiency in addition to regulatory compliance.

The research found that close to 90% of companies have little or no automation in place for the AI governance processes that they will need to rely on to ensure regulatory compliance, and fewer than 1 in 10 companies have fully or highly automated their model documentation process, which will make it challenging to comply with regulatory requirements for model transparency and explainability.

“AI regulations are a matter of ‘when,’ not ‘if.’ Leaders should view AI regulations as an opportunity to invest in tools that not only ensure compliance but also enable operational excellence in AI. First-movers that are better prepared for regulations also can build trust with customers, enhance their reputation and take market share from competitors that take a ‘go-slow’ approach to compliance,” Vartak said.