# Unfold CEO on Breaking Enterprise Data Barriers for AI Adoption Unfold extracts business context from legacy systems to enable effective AI agent operations. - **Source:** [The Next Web](https://brifing.si/source/afe24b20-b275-4c19-b9a6-8bb3c8babae7) - **Published:** 2026-11-08 - **Topic:** [technology](https://brifing.si/c/technology) - **Original article:** - **On Brifing:** https://brifing.si/s/c1301733-00bd-41da-a824-b08ba4131223 ## The Enterprise Data Bottleneck Enterprises face a critical obstacle as AI adoption accelerates: essential data remains trapped within legacy systems and proprietary software. Traditional integration tools often fail to capture the necessary business context or access the full scope of information required for AI. This limitation prevents organizations from fully leveraging their core operational data. ## Rethinking Data Access Strategies Unfold, led by CEO Ran Cherniavsky, addresses this by extracting information through existing interfaces without requiring disruptive system migrations. By mapping underlying logic directly from the UI and command lines, the platform bypasses the need for vendor-exposed APIs. This approach allows businesses to unlock data without replacing their foundational technology stacks. ## The Cost of Inaccessible Data Companies currently pay consultants to manually map system fields or rely on analysts to stitch together capped reports in Excel. These manual processes are slow, error-prone, and stall AI pilots because models lack the full picture. Consequently, organizations effectively rent access to their own data through restricted, vendor-controlled user interfaces. ## Modernization Risks and Realities Gartner reports that 45% of core systems still run on mainframes or on-premise client-server architectures. Furthermore, 34% of organizations now avoid modernizing core applications due to high risks, up from 16% in 2023. AI cannot wait for these long-term infrastructure overhauls, necessitating a solution that works with existing, older systems. ## Learning Systems Like Employees Unfold functions by learning each system similarly to a trained employee, identifying object meanings, business rules, and exceptions. By observing both the screen and the underlying data, the platform ties UI elements to their specific data locations. This process transforms raw, noisy data into a structured, meaningful catalog for AI agents. ## Deterministic Data Collection Process Onboarding any system takes up to 14 days, during which Unfold builds a deterministic, read-only collector. This collector operates through the system’s own permissions, ensuring consistent data retrieval. By mapping the full journey of every data object, the platform provides AI with both the raw information and the vital business context. ## Normalization Across Diverse Systems The platform maps disparate data points into a single canonical enterprise schema, ensuring consistency across different applications. For instance, it reconciles varying identifiers like KUNNR in SAP and Account in Salesforce. This normalization allows AI agents to receive consistent, unified data regardless of the source system's original naming conventions or structure. ## Feeding Modern AI Platforms Unfold acts as an upstream layer, feeding data into platforms like Snowflake, Databricks, or Splunk. It delivers context-rich data directly to agents on Claude or Gemini without requiring data movement. By providing this foundation, Unfold enables complex use cases, such as fraud detection using point-of-sale data from legacy AS/400 mainframe systems. ## Real-World Operational Impact Customers measure success through reduced time, lower consulting costs, and new operational capabilities. A global retailer successfully fed AS/400 point-of-sale data into a Gemini fraud-detection agent to satisfy regulatory requirements. Meanwhile, an auto dealership group enabled owners to query warranty money status in plain English, bypassing previously locked, report-based data silos. ## Infrastructure for the AI Era The current state of AI infrastructure mirrors the early days of the automobile, where powerful engines existed but lacked the necessary roads. Unfold aims to build these roads by providing the knowledge foundation for enterprise AI. This allows models to move beyond simple question-answering and begin running end-to-end business workflows. ## Future of Agentic AI Gartner projects the AI context platform market will grow from $28 billion in 2026 to $78 billion by 2030. As models become more intelligent, the primary challenge remains data access and control. Unfold positions itself to lead this category, ensuring that every enterprise AI use case runs on accurate, context-aware data. ## Sources covering this story - https://thenextweb.com/news/unfold-ran-cherniavsky-enterprise-data-ai-adoption — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._