Generative AI & Agents
From no-code to full-code — build and operate AI agents within governance.
Build and govern generative AI applications and AI agents at enterprise scale. It supports both no-code agent builders and full-code development, delivering a new class of AI apps that combine analytics, predictive models, and agents.
What Is Generative AI & Agents
Dataiku enables organizations to build and govern generative AI applications and AI agents at enterprise scale. Business users create agents directly with the no-code agent builder, while data scientists extend them with sophisticated full-code development. By combining analytics, predictive models, and agents, you build AI apps that go beyond simple chatbots to actually perform work — and every agent is centrally governed and evaluated.
Key Capabilities
No-code + full-code
Supports the no-code agent builder and full-code development side by side
Central agent control
Manage tools and permissions with Managed Agent Tools and Agent Connect
Continuous evaluation
Ongoing response-quality evaluation with Quality Guard (LLM-as-a-judge)
Internal knowledge integration
Grounded responses powered by RAG and Knowledge Bank
How You Build
| Audience | Approach |
|---|---|
| Business users & analysts | No-code agent builder |
| Data scientists & developers | Full-code (Python) development |
| Tools & integrations | Managed Agent Tools · Agent Connect |
| Quality & safety | Quality Guard · governance integration |
What It's Used For
Work automation agents
Agents that perform multi-step tasks such as document processing, lookup, and summarization
Internal knowledge assistants
RAG-based Q&A and research support applications
Analytics-infused apps
Decision-support apps combining predictive models and agents