🚀 New in Maia: AI Agent API endpoints, enhanced root cause analysis, and more!

Welcome back to the Maia New Features Blog! This week, we're excited to deliver updates that enhance your observability experience, make troubleshooting more intelligent, and unlock new programmatic possibilities with Maia. Let's explore what's new in our changelog updates.

🤖 Maia AI Agents' public API is here

We're excited to introduce the Agent Tasks API — Maia's first public API that lets you programmatically create, monitor, and manage tasks. This new capability allows you to send instructions to Maia AI Agents and have them executed as background processes. You now have the same level of control and agentic power via the API as you do when chatting to Maia AI Agents in Designer.

The API enables you to integrate Maia AI Agents’ problem-solving capabilities into your own workflows, CI/CD pipelines, or custom scripts—allowing Maia AI Agents to work through complex instructions as a background service.

What's new:

• From chat to API: Transition from manual 1-to-1 conversations to programmatic task creation. Define an objective for Maia AI Agents and manage the entire work lifecycle via the API.
• Asynchronous background tasks: Unlike a chat session that requires you to stay in the window, API tasks run independently. Send a request, receive a task ID, and check back whenever you're ready to review the output.
• Programmatic decision handling: Maia AI Agents still pause for safety when high-impact actions are needed. You can approve tool usage or answer clarifying questions through the dedicated endpoint to keep the task moving.
• Full Designer parity: You maintain the same level of granular control you have in the Designer—including the ability to manage permissions, track reasoning, and iterate on work—giving you complete command over Maia AI Agents' actions via the API.

How do I use it?

To get started, create a task by providing an initial instruction—this serves as your first prompt to Maia AI Agents. Once the task is created, you will receive a task ID which allows you to monitor the task’s status, handle decisions, and iterate on the task using follow-up messages.

Check out our API reference to try out the new endpoints, or read our guide to Using the Agent Tasks API for more details.

🔍 Root cause analysis gets smarter with enhanced pipeline visibility

Root Cause Analysis (RCA) has received significant improvements to make it more accurate and useful when pipeline failures occur. The enhanced RCA now provides comprehensive visibility across your entire pipeline infrastructure.

Key improvements include:

• Complete pipeline visibility: RCA now analyzes failures inside iterators and orchestration steps, not just top-level pipeline issues.
• Enhanced failure detection: Pipeline-level failures such as specific SQL errors are now captured and analyzed.
• Richer contextual analysis: Warning-level agent task logs are now included to provide the lead-up context needed to understand why failures occurred.
• Structured, actionable output: Every identified issue now includes a category and a clear indication of whether it's fixable within your pipeline.

For more information about Root Cause Analysis, check out our documentation.

💾 Saved views in Pipeline Run History

Pipeline Run History just got a major productivity boost! You can now save your filter sets and set a default view, eliminating the need to reapply filters every time you visit this page. Simply apply and save a view, and it’s there every time you need it.

💬 We'd love to hear from you!

Let us know how these new features are improving your workflows—we're all ears! Feel free to add any comments or questions below.

Want to get involved? Join the Matillion Community to stay up to date, share feedback, and help shape our product roadmap for future innovations.

‍