Bring AI agent data into your CX analytics with Data Hub

For many CX and data teams, AI agent interaction data lives in a silo, or stuck in manual exports. That gap makes it harder to connect automation to business outcomes.
You know the insights are essential, but it's just not where the rest of your analytics work happens.
That’s why we built Data Hub, a new capability in our AI agent management platform (AMP) that gives you secure, programmatic access to granular AI agent-customer interaction data that’s ready to query, enrich, and visualize in the analytics tools you already use. It brings event-level insights from every customer conversation with your AI agents straight into your analytic workflows, so you can track agent performance like any other critical business metric.
Whether you're trying to diagnose performance issues, measure CX outcomes, or simply understand the story behind the numbers, Data Hub puts your interaction data exactly where it belongs: in your stack, on your terms.
Streamlined access to all the agent interaction data you need
As suggested in CX Today, AI agents are only as effective as the data they can access. But in most contact centers, the context they need is in siloes across voice, chat, CRM, WEM, support. That fragmentation makes it harder for your teams to perform, and it shows up in clunky, inconsistent experiences for both customers and employees.
And even when you know this data is valuable, actually getting it out of a contact center platform is another story. It’s often buried behind manual exports, and by the time you get your hands on it, it’s too late to be useful.
Instead of delayed or limited access, Data Hub delivers the raw event stream directly—every message, escalation, and resolution, i.e., data where the personally identifiable information (PII) is redacted, structured, and refreshed on a predictable schedule so your team can work with it right away.
Here’s what’s included:
Built-in PII redaction for secure, compliant access out of the box
Raw event-level interaction data from every AMP conversation
Structured records, optimized for high-volume queries (partitioned by tenant and event month)
Automatic refreshes every 2 hours
Data access via Databricks Delta Sharing, and support for zero-copy queries
Built for the modern data analytics stack
You shouldn’t have to rethink your workflows to get insights into your AI agent performance. Data Hub integrates cleanly into the BI and analytics platforms you already rely on, so you can get to work without starting from scratch.
It works out of the box with:
Power BI, Tableau, Looker, and many more for reporting and dashboards
Analytic environments like BigQuery and Jupyter notebooks
Because it’s built on Delta Sharing, the data stays secure and governed, while still being easy to query without copying or moving it.
Track agent performance and the impact on CX
Summary metrics help you gauge performance at a glance, but they don’t tell the whole story. You need to know how your AI agent actually behaves, i.e., where it succeeds, where it stalls, and how it affects your bottom line. Well, now you can now get those insights.
Data Hub gives you the granularity and flexibility to:
Correlate AI and human agent behavior across the full support journey
When enriched with your own outcome data, such as CSAT, resolution rates, or churn, track the agent’s effect
Spot regressions or anomalies in near real time
Create and share metrics in dashboards your company already uses
Essentially, you can look at specific events—what the agent said, what tools they called, when the handoff happened—and line that up against what actually happened for the customer. This way, instead of guessing what went wrong or right, you can see it, measure it, and act on it.
Centralized support, seamless integration
Moreover, Data Hub is simple to set up and run. You manage access directly from AMP Studio, and once credentials are generated, you can connect and query right away.
You can:
Generate and rotate credentials from the AMP dashboard
Securely connect your BI tools or data environments via Delta Sharing
Easily verify the most recent data refresh using event timestamps
“With DataHub, our customers get instant access to rich, actionable data on AI agent performance—unlocking insights that drive insights-based optimization and faster iteration.
DataHub is built to be effortlessly flexible, so you can plug it into your stack and shape the data to fit your workflow and focus on what matters: Improve your AI Agent's performance.”
Elisabeth Reitmayr, Head of Engineering – AI & Data, Parloa
Because everything from credential setup to data streaming is self-service and handled in AMP Studio, you’re not filing tickets or waiting on integrations.
The foundation for serious CX analytics
Ultimately, if your team is already tracking customer experience, agent productivity, or operational efficiency, Data Hub helps you go deeper. It connects one of the most valuable (and often overlooked) CX data sources—AI agent interactions—into the reporting environments your company already knows.
Contact us to learn more.
Setting up Data Hub is fast and fully self-serve. Here’s how to get started:
Access the Credentials Dashboard Go to AMP Studio → DataHub Credentials to view and manage your credentials.
Create new credentials Click Create new, give your credential a clear name (e.g. marketing_dashboard_v1), and hit Create. Credentials are generated instantly and added to your list.
Retrieve your activation URL Use the three-dot menu to select Activation URL, then copy and securely store the download link for your .share credential file.
Connect via Delta Sharing Use the .share file to authenticate and access the data. You can connect using:
SDKs in Python, Java, JavaScript, or Spark
BI tools like Power BI, Tableau, or Looker
Manage and rotate credentials as needed From AMP Studio, you can view metadata, rotate tokens, or remove outdated credentials.
For detailed setup instructions and supported tools, check the Data Hub documentation.