The News
CData announced that its Managed MCP Platform, Connect AI, is now available on the Databricks Marketplace, enabling Databricks Agent Bricks to access live, contextual data from more than 350 enterprise systems. This partnership gives organizations a way to quickly build and deploy AI agents that can read, write, and act on both Databricks-native data and real-time operational data from sources such as NetSuite, Salesforce, SAP, and ServiceNow. To read more, visit the original announcement here.
Analysis
AI Agents Need Context, and the Market Is Struggling to Supply It
The enterprise AI landscape is shifting rapidly as organizations move from experimentation toward production-grade agents that automate workflows, make decisions, and operate on real business data. Our data shows that AI/ML spending continues to accelerate, with 70.4% of organizations prioritizing AI/ML tools in the next year. Yet progress is held back by an increasingly fragmented data environment. Companies now operate across hybrid and multi-cloud footprints, with 54.4% primarily hybrid and using an average of three to five cloud providers, according to Day 0 and Day 2 survey findings.
This fragmentation creates gaps in organizational context. Agents built on modern AI frameworks often have access to curated data within platforms like Databricks, but very limited visibility into operational systems where business processes actually live. Without a consistent source of truth for schemas, permissions, relationships, and semantics, AI agents struggle to execute meaningful actions reliably. These challenges echo a broader trend: developers want to build AI-native applications, but the underlying data access and governance layers remain the bottleneck.
A New Phase of AI Connectivity Emerges
CData’s partnership with Databricks lands at a critical moment in the ecosystem. The Model Context Protocol (MCP) is rapidly becoming the connective tissue for agentic architectures, offering a standardized way for AI agents to interact with external tools and data systems. By delivering a fully managed, production-ready MCP platform through Databricks Marketplace, CData brings a missing capability directly into the developer workflow: out-of-the-box, secure, real-time enterprise data connectivity that doesn’t require new pipelines or custom integration work.
For developers building with Agent Bricks, this approach may streamline the creation of context-rich agents that can reason across both analytical and operational data. Connect AI’s semantic intelligence layer, which interprets schemas and business logic from each connected system, aligns with the market’s move toward metadata-driven AI. theCUBE Research and ECI’s Day 1 data shows 57.2% of organizations rely on fully automated, reliable dependency and configuration management, signaling demand for platforms that can abstract complexity and provide standardized, machine-readable context.
This integration between Databricks’ Data Intelligence Platform and CData’s managed MCP capabilities may help reduce context fragmentation and improve reliability as agents move into production environments.
Current Market Challenges and Insights
Enterprises face intensifying pressure to operationalize AI while maintaining governance, compliance, and performance. The complexity is rising faster than teams can manage manually. Many organizations still rely on a mix of legacy integration middleware, custom pipelines, and brittle API connectors that limit scalability and slow down AI adoption. Day 0 data shows that 41.1% of teams still employ manual processes for configuration, and this manual overhead does not translate well into AI-driven environments where real-time context is required.
Security and governance complicate matters further. With 50.7% citing limited tools for securing infrastructure configurations and nearly half identifying vulnerabilities weekly, enterprises are cautious about exposing operational systems to autonomous agents. Integration platforms must enforce inherited permissions, identity controls, and compliance frameworks by default, which are requirements that most developer teams cannot shoulder alone.
At the same time, AI’s operational footprint is expanding. 84.5% of organizations already use AI for real-time issue detection, and 80.5% for performance optimization, signaling rising expectations for intelligent, context-aware automation across the business. The combination of fragmented data and rising autonomy demands a new class of architecture centered on standardized connectivity, semantic awareness, and secure execution.
These pressures explain the rising relevance of MCP and managed connectivity platforms: enterprises need ways to scale AI safely without rebuilding data access patterns system by system.
What This Means for Developers Moving Forward
While adoption levels will vary, CData’s Managed MCP Platform may influence how developers design, deploy, and govern AI agents in Databricks environments. By offering live, permissioned access to operational systems through a standardized protocol, developers may be able to reduce the number of custom connectors they maintain and shift toward more declarative, metadata-driven patterns. This can help shorten development cycles, improve the reliability of automated workflows, and simplify the creation of multi-step, multi-source agentic applications.
Real-time access to enterprise systems may also help developers build agents that respond to the actual state of the business rather than relying on batch data or static snapshots. This could support more dynamic decisioning, adaptive workflows, and higher-fidelity actions across systems, provided governance guardrails remain strong. Developers may also find it easier to enforce consistent identity and access controls when relying on platforms that inherit permissions and surface auditability features.
As AI agents evolve from copilots to autonomous actors, platforms like Databricks combined with managed MCP connectivity may offer developers a more predictable and secure foundation for delivering production-grade automation without recreating integration logic from scratch.
Looking Ahead
The market is moving toward a more standardized, interoperable approach to AI agent connectivity, where access to enterprise systems is mediated through unified protocols, semantic layers, and managed platforms. As MCP gains traction, organizations may begin consolidating scattered integration patterns and reducing the operational friction associated with building agents that operate across multiple systems.
CData’s launch on Databricks Marketplace positions it within a growing ecosystem focused on real-time, governed, context-rich AI operations. The combination of enterprise-grade connectivity, semantics-aware data access, and Databricks’ intelligence layer signals a broader trend: AI agents will increasingly require unified business context to operate safely and effectively. As platforms mature, expect deeper integrations, broader protocol adoption, and more robust guardrails designed to bring AI agents into production environments with confidence.
2026 Prediction: Cost-Aware Application Development Becomes a Core Engineering Discipline
2026 Predictions: Security Operations Become an AI-Augmented Workflow
2026 Predictions: Software Supply Chain Security Shifts to Continuous Verification
2026 Predictions: Application Security Modernizes Around AI, APIs, and Behavioral Controls
2026 Predictions: Platform Engineering Centralizes Security as a Service
