Enterprises accelerating AI adoption face a critical obstacle: data essential for AI often remains locked inside legacy platforms, proprietary software, and applications with limited interfaces. Ran Cherniavsky, CEO of Unfold, shares how his company’s breakthrough method tackles this challenge by accessing data through existing user interfaces, enabling AI to fully leverage enterprise information.
- Unfold extracts enterprise data via existing system interfaces without migrations.
- Legacy systems and partial APIs limit AI’s access to critical data and context.
- AI-driven data accessibility is key to security, modernization, and competitiveness.
What happened
Ran Cherniavsky, CEO and Co-Founder of Unfold, explains the persistent challenge enterprises face in making their data AI-ready. Much of the needed data remains locked inside legacy systems, proprietary applications, and platforms that offer only limited or partial API access. While traditional integration tools connect systems on a field-by-field basis, they fail to capture the full scope of information and business context required by AI technologies.
Unfold’s solution focuses on interacting with enterprise systems through the same interfaces employees use—whether user interfaces, reports, or command lines. This approach enables Unfold to rapidly map data locations, interpret underlying business logic, and extract meaningful, contextualized information without costly system rewrites or migrations.
Why it matters
AI demands comprehensive and context-rich data to function effectively, but existing integration methods only expose limited slices of data accessible through vendor APIs. This creates a bottleneck that slows AI pilots and reduces their accuracy, forcing companies to rely on manual data extraction and analysis. As AI adoption accelerates, enterprises unable to access full datasets risk falling behind competitors that leverage comprehensive AI insights.
Additionally, locked data poses security and modernization challenges. Organizations cannot secure what they cannot fully see or understand. With a significant portion of core enterprise systems still running on legacy mainframes or on-premises software, modernization remains risky and slow. Unfold’s approach enables AI readiness without waiting for costly modernization, addressing a critical pain point for enterprises pushing towards digital transformation.
What to watch next
As Unfold continues to develop its platform, the onboarding process promises to condense system integration timelines from months to as few as two weeks for any system with a user interface or command line. This scalability could accelerate AI adoption across industries using varied legacy infrastructures such as AS/400 mainframes, on-prem ERPs, and partial SaaS implementations.
Enterprises and AI vendors alike will be closely monitoring how effectively solutions like Unfold’s can deliver reliable, contextual, and comprehensive data access. Success here will not only enhance AI model accuracy and usability but will also set new standards for enterprise AI security and modernization, redefining competitive dynamics in digital transformation journeys.