The Fort Worth Press - GoodData.AI Launches the Agentic Serving Plane, a New Category of Enterprise Data Infrastructure for AI Agents

SAN FRANCISCO, CA / ACCESS Newswire / November 10, 2026 / GoodData.AI, the open agentic analytics platform, today launched the Agentic Serving Plane, a governed execution layer designed for the high-concurrency, context-rich data workloads created by AI agents.

The Agentic Serving Plane sits between enterprise data platforms and the applications, analytics experiences and agents that consume them. It provides workload serving, business semantics, contextual grounding, access control, and AI execution services through a common governed layer.

The current enterprise data infrastructure was designed primarily for humans: occasional queries, predictable workloads, and access by known users. AI agents operate differently. A single business task can generate tens or hundreds of short-lived data operations as an agent investigates, reasons, and decides what to ask next. When many agents operate simultaneously, intermittent human querying becomes continuous, high-concurrency agentic demand.

Agents also need more than raw data. They need business definitions, relationships, permissions, policies, and task-specific context on every request. Without a serving layer designed for those workloads, agent traffic can become slow, expensive, or disruptive to underlying data platforms.

At the same time, the enterprise data stack is becoming more fragmented. Cloud warehouses and lakehouse platforms increasingly provide their own compute, semantic and AI capabilities, while enterprises typically operate across multiple data platforms, applications and analytical environments.

The result is a new infrastructure requirement: a common layer that keeps business definitions consistent, isolates machine workloads, enforces access policies, and provides governed context regardless of where the underlying data resides.

For years, the market judged BI by the dashboard. Underneath it, GoodData.AI was building a governed system for serving data to applications: semantics, certified metrics, query planning, caching, permissions, and tenant isolation. With AI, those capabilities become the product.

Roman Stanek, CEO and Founder, GoodData.AI

Business Value for Enterprises

The Agentic Serving Plane supports the different ways people and systems consume enterprise data, from dashboards and embedded applications to assistants, copilots and autonomous agents.

Enterprises can use the same governed foundation to build internal analytics, operational AI applications, customer-facing data products and premium AI-powered services.

Agents can bring analysis directly into operational workflows, investigate business conditions across multiple steps, and increasingly automate follow-up actions. Applications and analytical experiences can reuse the same business definitions, permissions, and execution services.

Because each new application or agent uses the same underlying serving, semantic, contextual, and governance capabilities, teams do not have to rebuild the data foundation for every use case. This can shorten development cycles, reduce duplicated infrastructure, and give enterprise IT a common control layer across analytical and AI workloads.

How the Agentic Serving Plane Works

The Agentic Serving Plane is designed to work with the data infrastructure enterprises already operate.

Using open technologies, including Apache Iceberg and Apache Arrow, GoodData.AI can serve analytical and agent workloads from customer-controlled data environments while adding the semantic, contextual, governance, and execution capabilities required by people and AI systems.

Consider an agent investigating a decline in revenue. It may first compare regions, then examine product lines, then identify individual accounts, with each result determining the next query.

The Agentic Serving Plane manages that sequence as a governed workload. It applies the appropriate business definitions and access policies to every request, isolates agent traffic from other workloads, and is designed to maintain consistent serving performance as machine-generated query volume increases.

The architecture builds on the technology GoodData.AI has developed since its founding for software companies embedding analytics into their own products, enabling many customers to share a single platform while each tenant's data, permissions,, and business context remain isolated.

Gartner named GoodData.AI a Visionary in the 2026 Gartner^® Magic Quadrant^™ for Analytics and Business Intelligence Platforms, recognizing the company's broader analytics and developer-oriented platform approach.

Inside the Agentic Serving Plane

High-concurrency workload serving: A query architecture built on Apache Arrow and Apache Iceberg is designed to serve machine-generated analytical workloads efficiently while reducing pressure on production warehouse compute. Usage controls can be applied to individual users and agents.

Governed business semantics: Teams define metrics and business concepts once and reuse them across dashboards, applications and AI experiences, helping ensure that agents use the same definitions as established reporting and analytical workflows.

Delegated access control: Agents operate within explicit authorization boundaries. Permissions can be enforced at the field and row level alongside tenant isolation and other enterprise access policies.

Open agent integration: Through the GoodData.AI MCP Server, agents and AI tools can access the same governed metrics, analytical objects and business context used elsewhere across the platform.

AI observability and auditability: AI Observability records how responses are generated, giving data, AI, security, and compliance teams visibility into the data, context, and execution path behind an answer.

Context-aware execution: The serving plane combines data access with the semantic and operational context that agents need to determine what information they can use, which definitions apply, and when a request falls outside an authorized or defined boundary.

Native Inference Coming to the Agentic Serving Plane

GoodData.AI is also adding native inference capabilities to the Agentic Serving Plane.

Local Inference, planned to begin rolling out in the coming months, will allow selected AI capabilities to run closer to governed enterprise data without requiring a third-party model provider in the data path.

The architecture is designed to reduce latency and external inference cost while keeping AI execution closely aligned with customer-controlled business definitions, permissions and data context.

A person opens a dashboard and picks a filter. An agent keeps asking until the task is done, and its authority may need to be narrower than that of the person who initiated it. When a question cannot be answered within the available semantic, policy or data boundary, the system should say so rather than produce a plausible number.

Ondrej Macek, VP of Product, GoodData.AI

Stanek describes the strategy behind the Agentic Serving Plane in his blog post, Built for When the Agent Asks Next.

About GoodData.AI

GoodData.AI is an open agentic analytics platform that helps enterprises put AI to work on their data while maintaining control over business definitions, permissions and execution context.

Its agents can follow through on multi-step business processes rather than answering a single question and stopping. Their analysis and actions are grounded in customer-defined semantics, policies and context so that organizations retain control over how enterprise data is interpreted and used.

GoodData.AI's Agentic Serving Plane provides a governed execution layer between enterprise data and the dashboards, applications, assistants and agents that consume it. Enterprises can use the same foundation to support multiple analytical and AI experiences without rebuilding governance for each one.

The platform supports customer-controlled infrastructure, bring-your-own-LLM flexibility, MCP and A2A integration, and open development through APIs and SDKs.

Headquartered in San Francisco with engineering based in Prague, GoodData.AI serves enterprises and software companies worldwide.

For more information, visit GoodData.AI and follow GoodData on LinkedIn, YouTube, and Medium.

Contact:
Natalia Nanistova
[email protected]

SOURCE: GoodData.AI

View the original press release on ACCESS Newswire

J.Barnes--TFWP