Self-service analytics promises faster answers, yet many organizations end up with conflicting dashboards, duplicated or leaky data, and renewed reliance on IT.
The issue is not control versus flexibility, its delivering governed, analytics-ready data that supports secure independent exploration and consistent reporting.
DBTA recently held a webinar, Self-Service Analytics Without the Chaos: Governing Insights at Scale, with Vaibhav Suresh, product marketing senior analyst at Informatica and Simon Thornell, field CTO at TrustLogix, who explored the architectural and operational practices required to deliver both flexibility and trust.
Suresh opened his conversation by citing a stat from “The State of Data and Analytics Report 2025,” 70% of data leaders believe their most valuable business insights reside within data that is siloed, inaccessible, or otherwise unusable.
Finding trusted data can be challenging, he explained. Disparate sources and methods to access data create risks for the organization. It’s nearly impossible to know who has access to what, to manage compliance goals and to optimize data delivery.
According to Suresh, when preparing for the agentic enterprise, agents need:
- Scale: Massive volumes of verified data products
- Speed: Real-time accessibility to trusted data
- Context: Explore the lineage, quality, and meaning of data
- Safety: Provide data protection at enterprise scale
The four cornerstones of successful data sharing include finding, understanding, trusting, and accessing, Suresh said.
Workers need to be able to locate relevant data and Metadata. Context is king. Companies need to know where the data originated, who owns it, who else has used it, who has identified it as a solid source, was it used successfully, etc. The data needs to be high-quality and trustworthy. And people within the organization need to be able to have it prepared and delivered upon request or approval.
There are four areas reshaping self-service analytics, Thornell noted.
Democratized creation: Business users and AI assistants now author more analytical content than central BI teams. The producer/consumer line has blurred.
Natural language is the new way to query: Based on intent SQL is now generated, not authored. Policies must enforce against queries no human ever wrote or reviewed.
Semantic layers becoming the single version of truth: Governed metrics live in the semantic layer—but row level access often doesn’t. Now a policy problem, not just a definition problem.
Open formats and data sharing: The same Iceberg or Delta table is read by Snowflake, Databricks, Trino, DuckDB and shared across companies via marketplaces. One policy must travel with the data.
This is where TrustLogix solutions come into play, Thornell said. TrustLogix maintains row-level and column level controls into PowerBI… Including imported datasets by mapping source identity and policy through to RLS in Power BI. TrustLogix monitors access risks from both users and agents, applies attribute-based policies at MCP servers and AI tool calls, and enforces those same controls natively in Snowflake, Databricks, and the rest of your data platforms. The TrustLogix platform offers one control plane for data and AI.
For the full webinar, featuring a more in-depth discussion, Q&A, and more, you can view an archived version of the webinar here.