GoodData.AI
by GoodData Corporation
Governed semantic layer and serving plane for AI agents and embedded analytics
GoodData.AI is an analytics and data platform that lets enterprises and software vendors serve governed metrics to dashboards, applications and AI agents from one semantic layer. It is for data and platform teams that need agents to query business data with consistent definitions, row-level permissions and an audit trail, without overloading the warehouse.
GoodData.AI, renamed from GoodData in April 2026, is a San Francisco analytics vendor founded in 2007 by CEO Roman Stanek, with engineering in Prague. Its platform centres on a governed semantic layer where metrics are defined once (historically in its MAQL language) and reused by dashboards, embedded analytics and AI. Through 2026 it added an MCP Server in January that lets agents build and operate analytics end to end, Agent Builder in April for code-first or no-code agents with guardrails, tracing and Agent2Agent support, and AI Observability in September for interaction-level tracing. On October 6, 2026 it launched the Agentic Serving Plane, a governed execution layer between data platforms and the agents and apps that consume them. It combines high-concurrency serving built on Apache Arrow and Apache Iceberg to keep machine-generated query volume off warehouse compute, per-user and per-agent usage controls, delegated field- and row-level access with tenant isolation, and a record of the data and context behind each answer. Local inference, which would run selected AI features without a third-party model in the data path, is promised for the coming months. The company says it serves over 123,000 companies and 3.9 million users, many through software vendors embedding its analytics, and Gartner named it a Visionary in the 2026 Magic Quadrant for Analytics and BI Platforms. Pricing is per workspace rather than per user, which suits multi-tenant ISVs.
A head of data or platform engineering at an enterprise or ISV that already serves metrics to many tenants and now needs AI agents to query the same governed definitions safely.
Agents, dashboards and embedded apps answer from one set of governed metrics and permissions, with usage limits that keep agent query volume from driving up warehouse cost.
At a Glance
- Category
- Data & Analytics
- Pricing
- Subscription, Usage-based, Contact for pricing
- Target Market
- CTOs, Heads of Data, Data Engineers, ISV Product Leaders, Enterprise Developers
- Deployment
- Cloud-first, Self-hosted, Multi-cloud
- Founded
- 2007
- Headquarters
- San Francisco, United States
- Customers
- 123,000+ companies and 3.9M users (company claim, April 2026)
Key Features
- ✓Agentic Serving Plane
A governed execution layer between data platforms and the dashboards, apps and agents that consume data, launched October 6, 2026.
- ✓High-concurrency serving on Arrow and Iceberg
Absorbs machine-generated agent query volume and applies per-user and per-agent limits, reducing load on production warehouse compute.
- ✓Governed semantic layer
Metrics and business concepts are defined once and reused across dashboards, embedded apps and AI, so agents and humans get the same numbers.
- ✓Delegated access control
Agents act within explicit authorisation limits using field- and row-level permissions and tenant isolation inherited from the platform.
- ✓MCP Server and Agent Builder
External agents reach governed metrics over MCP, and teams can build their own agents with guardrails, tracing and Agent2Agent support.
- ✓AI Observability
Records the data, context and execution path behind each AI answer, giving compliance and engineering teams a trace to audit.
- ✓Flexible deployment
Managed SaaS on AWS or Azure in US, EU and Australia, with dedicated, multi-region and self-hosted options on Enterprise.
Capabilities
Use Cases
- •AI assistant inside a SaaS product
An ISV gives each customer tenant an agent that answers questions from that tenant's governed metrics, with isolation enforced by the platform.
- •Governed data access for internal agents
An enterprise points its agents at GoodData.AI over MCP so they query approved metric definitions instead of writing ad hoc SQL against the warehouse.
- •Controlling warehouse cost from agent traffic
Platform teams route high-volume agent queries through the serving layer and cap usage per agent so compute bills stay predictable.
- •Auditable AI answers in regulated industries
Compliance teams review the recorded data and execution path behind an AI-generated figure before it is used in reporting.
- •BI modernisation
Companies migrate legacy dashboards onto a code-defined semantic layer that then serves both people and AI from the same model.
Ideal For
Best For
- ✓Software vendors embedding multi-tenant analytics and AI assistants in their own product
- ✓Enterprises that want AI agents to query governed metrics instead of raw warehouse tables
- ✓Regulated companies that need row-level permissions, tenant isolation and an audit trail for AI answers
- ✓Teams that need self-hosted or customer-controlled deployment with bring-your-own-LLM
- ✓Data teams exposing a semantic layer to external agents over MCP
Not Ideal For
- ✗Business users who want a quick self-serve BI tool: reviewers consistently report a steep learning curve around MAQL and data modelling
- ✗Small teams wanting transparent pricing, since every plan is quote-only on annual contracts
- ✗Buyers who need the promised local inference today; it was announced as rolling out in the coming months
- ✗Organisations whose agent problem is data quality, which analysts say still has to be handled in the warehouse or an observability tool
Integrations
Deployment
Market & Ratings
123,000+ companies and 3.9M users (company claim, April 2026)
Market Analysis
Pros
- ✓Mature multi-tenant semantic layer with nearly two decades of embedded-analytics use behind it
- ✓Agent access is governed by the same row-level permissions and tenant isolation as dashboards
- ✓Deployment range from managed SaaS to self-hosted, with HIPAA and FedRAMP available on Enterprise
- ✓Analysts credit the semantic layer as a strength, and Gartner named it a Visionary in the 2026 BI Magic Quadrant
Cons
- ✗Steep learning curve: reviewers report MAQL is hard to learn and requires unlearning SQL habits
- ✗Constellation Research's Michael Ni said GoodData is not making agents easy to build
- ✗Data quality problems such as freshness and schema changes still have to be handled outside the platform, per Moor Insights
- ✗Independent reviews of the October 2026 serving plane are not yet available; claimed speed-ups are vendor figures
Pricing
Professional
Contact for pricing
- ✓Platform fee plus per-workspace pricing
- ✓Unlimited users and data
- ✓Managed SaaS on AWS or Azure (US, EU, Australia)
- ✓SOC 2, GDPR, ISO 27001
Enterprise
Contact for pricing
- ✓Custom use-case-based pricing
- ✓Dedicated cluster, multi-region or self-hosted
- ✓SAML and OIDC SSO, audit log
- ✓HIPAA and FedRAMP on demand
- ✓99.5% uptime SLA with 24/7 support
No list prices are published. Professional charges a platform fee plus the number of workspaces with unlimited users and data; Enterprise is custom. AI usage is metered by AI queries, with a default fair-usage cap of 30 per user per day and extra query buckets sold separately. Contracts are annual and cannot be downgraded mid-term. Per-workspace pricing contrasts with the per-user model most analytics vendors use.
Security & Compliance
Connect
Sources
This page was written from 7 sources, 4 on domains other than gooddata.ai.
- 1.gooddata.ai — gooddata ai launches the agentic serving plane a new categorvendor
- 2.gooddata.ai — pricingvendor
- 3.gooddata.ai — gooddata kicks off 2026 with mcp server launch and major entvendor
- 4.techtarget.com — GoodData joins agentic AI development mix with Agent Builder
- 5.peerspot.com — gooddata reviews
- 6.stockanalysis.com — gooddata
- 7.github.com — gooddata python sdk
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
Hex
Agentic analytics platform: AI notebooks for data teams and conversational Threads for everyone, grounded in governed semantic context
Unstructured
ETL for unstructured data: turn PDFs, tables, emails and scans into AI-ready JSON for RAG and agents
Tacnode Context Lake
PostgreSQL-compatible context lake giving AI agents shared, live, transactional state at decision time
Reducto
Agentic document parsing API that turns complex PDFs, scans and spreadsheets into citation-grounded data for AI, now with the 1¢-per-page r-1 model