Dataiku Launches AI Agents for Enterprise-Scale AI Applications

Dataiku has launched AI Agents with Dataiku, enabling companies to create and control AI agents at scale. The platform aims to provide a new class of AI applications powered by analytics, predictive models, and agents, addressing the issues of uncontrolled and ungoverned agent usage. Dataiku's capabilities include visual agent creation, code agent creation, managed agent tools, GenAI registry, sign-offs for risk monitoring, and a Dataiku LLM Mesh to manage model access.

Artificial intelligence data analytics startup Dataiku Inc. has announced the launch of AI Agents with Dataiku, a new set of capabilities designed to create and control AI agents at scale. This offering addresses the challenges faced by companies rushing to deploy AI agents, often resulting in clumsy architectures with varying quality and relevance, and sprawled ungoverned across teams [1].

The new platform aims to deliver AI applications powered by analytics, predictive models, and agents. Dataiku's Universal AI Platform is augmented with AI agents as true enterprise systems, grounded in trusted data, embedded in operational workflows, and governed with the same rigor as any business-critical asset. The platform introduces two development tracks: a visual agent for nontechnical users and a code agent for developers, allowing cross-functional teams to collaborate without sacrificing control or flexibility [1].

Dataiku has implemented the "LLM Mesh" architecture to manage access to models from providers like OpenAI, Anthropic, Mistral, and open-source options such as Llama. This supports both self-hosted and cloud-based deployments across services like AWS Bedrock, Microsoft Azure, and Google Gemini, allowing companies to integrate AI agents into their preferred infrastructure while maintaining control over data residency and access [1].

For security and governance, Dataiku introduces Dataiku Safe Guard, a feature that lets IT teams apply flexible guardrails across agent interactions. Combined with the Agent Connect feature, organizations can centralize agent usage and dispatch requests to single or multiple agents as needed. Performance and observability tools like Trace Explorer and Quality Guard ensure transparency into agent decision-making and evaluate outputs [1].

Central to the new offering is the generative AI Registry, allowing business leaders to review agent use cases, assess risk, and determine business value before agents are moved into production. The centralized registry supports broader efforts to align agent deployment with organizational priorities [1].

Dataiku co-founder and Chief Executive Florian Douetteau stated, "As companies mature from exploring to operationalizing agents, they’ll quickly realize the need to centralize their creation for proper governance instead of letting them sprawl, continuously optimize their performance instead of letting them stale and fully orchestrate them through existing IT assets instead of raising technical debt" [1].

As the number of AI agents continues to grow, so does the need for robust cybersecurity measures. SandboxAQ, a leader in AI and cybersecurity solutions, has announced AQtive Guard, a platform designed to manage and secure Non-Human Identities (NHIs) and cryptographic assets used by AI agents. The platform addresses the evolving cyber threats posed by AI agents, enabling real-time vulnerability detection and mitigation [2].

AQtive Guard offers technical integrations with CrowdStrike and Palo Alto Networks, providing organizations with full visibility into their non-human identity and cryptographic inventories and vulnerabilities. The platform uses Large Quantitative Models (LQMs) to deliver AI-powered insights, prioritization, and risk analysis, as well as automated remediation and lifecycle management [2].

With the rise of AI agents, the need for comprehensive governance and security measures becomes increasingly critical. Dataiku's AI Agents with Dataiku and SandboxAQ's AQtive Guard represent significant advancements in addressing these challenges, offering companies the tools needed to deploy and manage AI agents securely and effectively.

References:
[1] https://siliconangle.com/2025/04/24/dataiku-introduces-ai-agents-unify-govern-enterprise-agent-deployments/
[2] https://www.prnewswire.com/news-releases/more-than-1-billion-ai-agents-spawn-a-vast-new-cyber-attack-surface-sandboxaq-launches-new-platform-to-address-the-threat-302434143.html

Ask Aime: How do companies navigate the challenges of deploying AI agents effectively?

View Source