Logo

Dataiku Agent Management

Know if your AI agents actually deliver value. Discover AI agents across platforms and see the value each one brings to your business.

Request a demo

Govern at scale

Monitor your entire agent landscape.

Manage AI agents with full visibility and control

Discover every agent, wherever it was built

Scan agents across platforms into one inventory, each agent with an owner and a purpose.

See whether each agent is working

Track metrics like health, active users, and cost for every agent. Get alerts when a metric goes off track.

Control the risk, on the record

Certify agents and test them on a schedule. Stop an agent from one place, and keep the risk record ready for auditors.

One place for every agent, across every platform

Companies run agents across many platforms with no complete, up-to-date list of what exists. Agent Management scans those platforms into one inventory so every team can see what it's running, who owns it, and what it's for.

Agent Management - One place for every agent, across every platform

Watch how the agent behaves, not just if it’s running

Infrastructure monitoring tells you if an agent is responding, not if it’s doing its job. Agent Management tracks each agent’s health, active users, and token cost against your KPIs, so you see cost and value side by side.

Agent management - Watch how the agent behaves, not just if it is running

When an agent carries real risk, keep the record

Certify agents so your teams can trust and reuse them. Check each one against named risks, re-run tests on a schedule, and stop an agent from one place. When an auditor asks, the record’s already there.

Agent Management When an agent carries real risk, keep the record

Oversee agents from the platforms your teams are using

AWS logo svg
Databricks logo svg
Google cloud logo svg
Azure logo svg
Microsoft Logo
Salesforce.com logo
Snowflake

Loved by customers and recognized by analysts

Dataiku named a Leader in the 2026 Gartner® Magic Quadrant™

For the fifth consecutive year, Dataiku has been named a Leader in the Gartner® Magic Quadrant™ for AI Platforms for Data Science and ML, once again recognized for both "Completeness of Vision" and "Ability to Execute."

“The platform is intuitive, collaborative, and streamlines workflows from data prep to model deployment. Dataiku has truly transformed how we handle data!”

Data scientist

Retail

Explore Dataiku agentic AI features

Dataiku Cobuild

The natural-language way to work across Dataiku. Get trusted answers, build governed AI, and oversee every agent you run.

Explore Dataiku Cobuild

Dataiku Agent Hub

Scale AI agents with full control by creating, connecting, and operationalizing them within governed workflows.

Explore Agent Hub

Dataiku E2A (Expert-to-Agent)

Scale expertise with production-ready AI agents grounded in enterprise data, structured reasoning, and human oversight.

Explore E2A

Explore Agent Management stories

Watch the web series: Governing AI Agents

AI agents are moving from prototypes to production, but only when they’re reliable, safe, and well-governed. In this webinar series

Watch the webinars

Resource thumbnail

Every enterprise is building agents. Almost none can account for them.

Resource thumbnail

Why traditional observability misses AI agent failure

Resource thumbnail

The invisible fleet: why agent sprawl is now a compliance liability

Request a demo of Dataiku

Get introduced to the latest features of the platform for AI success and explore what AI success will look like for you.

Frequently asked questions

Dataiku Agent Management is a standalone product that gives an organization one place to discover, monitor, and manage the risk of all its AI agents, wherever they were built.
Dataiku Agent Management connects to the platforms where teams create agents: Microsoft Copilot Studio and Azure Foundry, Salesforce Agentforce, AWS Bedrock, Google Gemini Enterprise Agent Platform, Databricks, Snowflake Cortex, and Dataiku itself. OpenTelemetry support extends coverage to agents built on custom and open-source frameworks.
No. A gateway sits between agents and their tools and controls traffic. Agent Management sits above the agents and measures their value for the company. For agents built in Dataiku, the LLM Mesh already plays the gateway role, and Agent Management sits on top of that too.
Agent Management connects to third-party frameworks through native APIs and log streams. Users simply select the framework, authenticate, and the system automatically scans and inventories the agents. The depth of integration depends on what each platform exposes, with deeper integrations actively being developed.
Yes. Stopping an agent from Agent Management is supported on select platforms. Agent Management isn’t in the traffic path between agents and users, so runtime controls stay with each agent’s platform.
Yes. Agent Management counts each agent’s daily, weekly, and monthly active users from its actual traces, and converts its token consumption into cost using a rate table you can adjust to your own pricing. Cost can be viewed per agent or rolled up across groups of agents, such as a business unit or program.
Agent Management connects to each platform through its native APIs, using credentials set up by your administrators. It reads agent metadata to build the inventory and, for monitored agents, collects their traces to compute metrics like usage, cost, and performance. Traces are stored in your Agent Management instance, accessible only to administrators, and conversations are not displayed in the product interface. Permissions control which users can see each agent and its metrics.