Dataiku
Also known as: Dataiku Universal AI Platform, LLM Mesh
Enterprise AI and machine learning platform with an agentic layer for building and governing agents.
Dataiku is an enterprise AI platform that has extended its data preparation, machine learning, and analytics environment to build, run, and govern AI agents in the same place. The company was founded in 2013 and has offices in New York and Paris. Analysts, data scientists, and developers work side by side through no-code, low-code, and full-code interfaces on one governed foundation.
Agents come in several forms: simple visual agents, structured visual agents that lay out a sequence with routing, delegation to other agents, parallel branches, loops, and reflection, code agents in Python, and external agents from Snowflake Cortex, Databricks, AWS Bedrock, and Google Vertex AI.
They call tools such as API endpoints, OpenAPI specifications, MCP servers, and Knowledge Bank search over vector stores with document-level security, can keep long-term memory per user, and can be set to pause for a person's approval before a sensitive tool call.
Underneath sits the LLM Mesh, which gives administrators governed connections to OpenAI, Anthropic, Bedrock, Vertex, Mistral, NVIDIA NIM, Snowflake Cortex, locally hosted models, and others, with cost control, rate limiting, and guardrails for PII, prompt injection, and toxicity.
Agents reach people through Agent Hub, Agent Chat, Slack, and Microsoft Teams, run unattended inside automation scenarios on schedules or data changes, and are exposed to other systems through REST and Python APIs, an A2A server, and an MCP server. Every call returns a nested trace with tokens and cost for the Trace Explorer, and an Evaluate Agent recipe and Agent Review score answers and tool paths across versions. Dataiku Solutions add ready-made GenAI and agent solutions for selling, supplier management, clinical trials, medical entity extraction, and finance insights.
Dataiku runs as the fully managed Dataiku Cloud or on a company's own infrastructure. Dataiku holds ISO 27001, ISO 27701, ISO 42001 and ISO 9001 certification, with SOC 1 and SOC 2 Type II reports for Dataiku Cloud and a HIPAA report. A free edition and a 14-day cloud trial are available, and paid editions are quoted by sales.
Vendor details
Canonical URL
https://dataiku.com
Category
Multi-agent platform
Subcategory
Enterprise AI and ML platform with agentic layer
Funding status
Privately held enterprise AI company founded in 2013, with headquarters in New York and Paris.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Connects to any cloud provider, data platform, and GenAI service, with the LLM Mesh routing across more than a dozen model providers including OpenAI, Anthropic, and Mistral. A plugin ecosystem, full Python and LangChain support, and Agent Hub let teams integrate external agents, tools, and enterprise data into governed workflows.
In practice
You want agents but refuse to lock into one model vendor. Dataiku's LLM Mesh routes agents across OpenAI, Anthropic, Mistral, and self hosted models, so you pick the provider per use case.
Compliance will not let unmonitored agents touch production. Dataiku tracks agents in the same central registry as datasets and models, blocks deployment until the required approvals are signed off, traces each tool call, and tracks agent performance and drift after deployment.
Analysts and data scientists need to build agents on the same trusted data. Dataiku gives a no code canvas and full Python with LangChain over one governed layer, grounding agents in your pipelines and business logic.
Sources & related URLs
Research sources
Agentic Index coverage score
13.0 / 14 capabilities · 93%
| Integrations & Tool Calling | Full |
|---|---|
|
Agents call managed tools, API endpoints, OpenAPI specs, local and remote MCP servers, Knowledge Bank search, LLM Mesh queries, web search, inline code and plugin tools, over connections secured per group. Sourcedoc.dataiku.com/dss/latest/agents/index.htmlread 2026-09-27 |
|
| Workflow Orchestration | Full |
|
Structured visual agents define sequences with routing, delegation to other agents, parallel and for each branches, agentic loops and reflection blocks, alongside code agents and external agents from Snowflake Cortex, Databricks, AWS Bedrock and Google Vertex AI; Agent Hub can orchestrate a query across several agents. Sourcedoc.dataiku.com/dss/latest/agents/structured-visual-agents/index.htmlread 2026-09-27 |
|
| Knowledge Grounding & RAG | Full |
|
Knowledge Banks embed datasets and documents into vector stores that agents search through a Knowledge Bank Search tool with static, dynamic and agent inferred filters and document level security, with GraphRAG and automated RAG optimization also documented. Sourcedoc.dataiku.com/dss/latest/agents/tools/knowledge-bank-search.htmlread 2026-09-27 |
|
| Human Oversight & Guardrails | Full |
|
Human approval set on a tool makes a visual agent pause before calling it and ask the user to confirm, optionally letting the person reviewing edit the tool inputs (not inside For Each, Parallel or Reflection sub sequences), and guardrails cover PII, prompt injection, toxicity, topic boundaries and bias. Sourcedoc.dataiku.com/dss/latest/agents/tools/human-approval.htmlread 2026-09-27 |
|
| Security, Identity & Governance | Full |
|
Dataiku offers SSO through SAML, OIDC, LDAP and Kerberos, MFA, user profiles, per project group and user permissions, connection security, user secrets and an audit trail. Its trust page lists ISO 27001:2022, ISO 27701, ISO 42001 and ISO 9001 certification, SOC 1 and SOC 2 Type II for Dataiku Cloud, and a HIPAA report. Sourcedataiku.com/legal/trustread 2026-09-27 |
|
| Observability & Auditability | Full |
|
Every LLM Mesh call returns a nested trace of the agent's LLM calls, guardrail steps, inputs, outputs, tokens and cost, viewed in the Trace Explorer (tree, timeline and explorer views), and the evaluation trajectory explorer replays tool calls in order; an audit trail records user actions. Sourcedoc.dataiku.com/dss/latest/agents/tracing.htmlread 2026-09-27 |
|
| Memory & State Persistence | Full |
|
Long term memory blocks store facts, such as preferences, and episodes from past conversations in memory banks, kept separately per user, shared by agents in a project through a default bank or split into named banks, and retrieved by LLM filtering and semantic search in later conversations. No retention period or delete path is stated. Sourcedoc.dataiku.com/dss/latest/agents/structured-visual-agents/blocks/long-term-memory.htmlread 2026-09-27 |
|
| Deployment & Data Residency | Full |
|
Dataiku runs as fully managed Dataiku Cloud or is installed on the customer's own infrastructure (Linux, macOS, experimental Windows) with hybrid options through sales, with installation, cloud and container deployment guides; locally running Hugging Face models keep inference in house. Sourcedataiku.com/product/get-startedread 2026-09-27 |
|
| Prebuilt Agents, Templates & Packs | Full |
|
Dataiku Solutions include named GenAI and agent solutions for separate jobs, Dynamic Selling Assistant, Supplier Management Assistant, Clinical Trial Intelligence Assistant, Medical Entity Extraction Assistant and Agentic Insights; Agent Hub distributes an organization's own governed agents and prompt library. Sourceknowledge.dataiku.com/latest/solutions/index.htmlread 2026-09-27 |
|
| Triggers & Channel Coverage | Full |
|
Agents run inside automation scenarios, which start on time based, dataset change, SQL query change, scenario completion or Python triggers without a person, and users reach agents in Slack (DM or mention), Microsoft Teams, Agent Hub, Agent Chat and through APIs. Sourcedoc.dataiku.com/dss/latest/scenarios/index.htmlread 2026-09-27 |
|
| Model Flexibility & Routing | Full |
|
Administrators define LLM Mesh connections to OpenAI, Azure OpenAI, Anthropic, AWS Bedrock and SageMaker, Microsoft Foundry, Google Vertex, Mistral, Cohere, NVIDIA NIM, Snowflake Cortex, Databricks foundation model APIs and locally running Hugging Face models, and agents pick from them, with cost control and rate limiting documented. Sourcedoc.dataiku.com/dss/latest/generative-ai/llm-connections.htmlread 2026-09-27 |
|
| APIs, SDKs & MCP Extensibility | Full |
|
Agents are exposed as A2A servers at a per agent public API endpoint with well known URI discovery and bearer API key auth, in JSON-RPC and SSE modes, beside a documented MCP server, the public REST API, the Python API and LangChain integration on developer.dataiku.com, and a plugin system. Sourcedoc.dataiku.com/dss/latest/agents/a2a.htmlread 2026-09-27 |
|
| Testing, Debugging & Optimization | Full |
|
The Evaluate Agent recipe scores answers and the path taken, with ordered and unordered tool call checks, RAGAS based LLM judges, BERTScore and custom metrics, stores results for side by side version comparison, and Agent Review lets builders and SMEs run test cases against an agent version with human verdicts overriding the judge. Sourcedoc.dataiku.com/dss/latest/agents/evaluation.htmlread 2026-09-27 |
|
| Browser & Computer Use | Not documented |
|
The agent tool list covers APIs, MCP, knowledge search, code and an LLM side web search tool, with no browser or computer use tool documented. Sourcedoc.dataiku.com/dss/latest/agents/index.htmlread 2026-09-27 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Product updates
Headless is a new plugin and MCP server that lets coding agents such as Claude Code, Codex and Cursor create and change Dataiku projects, including pipelines, models, dashboards and agents, by handing the build to Cobuild. It can also help move workflows built in SAS, Alteryx or Excel into Dataiku, and gives administrators tools that mainly inspect projects and failing jobs. It installs from the public Claude Code and Codex plugin marketplaces.
Bears on: MCP / tool calling / API
View sourceDSS 15.1.0 adds a Plan mode to Cobuild, Dataiku's natural language build agent, which proposes a set of high level tasks that users review and refine before the build starts. Once the plan is approved, Cobuild switches to Build mode and tracks progress against it. The same release adds an activity monitoring screen for deployed agents and a Secrets Manager for credentials used in connections and code.
Bears on: Agent capability
View sourceDataiku launched Cobuild on Snowflake, which combines Snowflake Cortex AI's model access with Dataiku's orchestration layer so users can build and validate AI workflows and agents in plain language.
Bears on: Integrations
View sourcePricing
Free Edition and a 14 day cloud trial. Paid editions are quoted by sales.
Paid editions are priced by number of users and by the platform capabilities enabled.
Included quota
The trial includes 2 users, 1 API service, 4 CPUs and 32 GiB of elastic compute for 14 days. The Free Edition installs locally, and its limits are not stated.
What is public
The Free Edition download and the 14 day Dataiku Cloud trial with its limits are public. Paid editions, hybrid deployment and commercial terms are quoted by sales. No paid price is published, and there is no pricing page.
Billing mechanics
Free Edition to start, then enterprise editions priced by number of users and enabled capabilities, sold through sales. Agentic features and the LLM Mesh are part of the platform. The Mesh adds central control over underlying LLM token spend.
Cost watchouts
Beyond platform licensing, running agents incurs LLM token costs from the chosen providers. The LLM Mesh Cost Guard helps enforce budgets, but heavy agent usage still drives model spend.
Variable cost rationale
Two things are metered. The license scales with the number of users and the capabilities enabled, and agents add usage based cost through the LLM tokens they consume from the chosen model providers, so the bill grows as more people build and as agents run more often. The LLM Mesh centralizes model spend, and its Cost Guard enforces budgets on it. The Free Edition runs on the customer's own machine at no license cost.
Additional watchouts
Enterprise pricing scales with users and capabilities and is not public. Separately, agents incur underlying LLM token costs, which the LLM Mesh helps cap but does not remove.
Sales call required
Yes, required for paid access
Free / trial
Free Edition to install locally on macOS, Linux or Windows (experimental). A 14 day Dataiku Cloud trial includes 2 users, 1 API service, 4 CPUs and 32 GiB of elastic compute, with no credit card and every feature except Govern and advanced LLM Mesh.
Lowest paid plan
Not published
Key ambiguities
No paid figure is published. Editions, users and deployment are quoted by sales, and model token costs sit on top.
Related vendors
- CrewAI — Open source multi agent framework (Python, MIT) of role based Crews…
- Relevance AI — Platform for building and deploying a workforce of AI agents with a…
- Agent Zero — Open-source multi-agent framework that gives its agent a full Linux…
- Akka — Enterprise agentic AI platform that builds and runs multi agent…
- Artian — Enterprise multi agent AI platform for financial services, where a…
- Databricks Agent Bricks — Databricks' agent platform, now sold as Agent Bricks, for building,…
Alternatives to Dataiku
The closest documented capability profiles to Dataiku among multi-agent platforms tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Databricks Agent Bricks13.0 / 14Matches Dataiku across all 14 documented capabilitiesDataiku vs Databricks Agent Bricks →
- Relevance AI13.0 / 14Matches Dataiku across all 14 documented capabilities
- Akka12.5 / 14A lighter documented profile than Dataiku
- CrewAI13.0 / 14Adds documented Browser & Computer Use
- Legion Intelligence11.5 / 14A lighter documented profile than Dataiku
- Rasa11.5 / 14A lighter documented profile than Dataiku
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded