The rise of Agentic BI is accelerating thanks to open source solutions and standards. With players such as Dremio, dbt, and Snowflake, companies are building data intelligence that is more sovereign, interoperable, and transparent.
Artificial Intelligence is accelerating the transformation of data-related professions through a new trend: Agentic BI, where intelligent agents can autonomously query, analyze, and interpret data.
However, this promise only makes sense if it relies on solid, governed, and interoperable foundations within an open ecosystem.
Open Source has become the lever that enables the convergence of initiatives and standards — as illustrated recently by dbt Labs confirming its commitment to the Open Semantic Interchange (OSI) initiative led by Snowflake, by releasing its semantic engine MetricFlow under the Apache 2.0 license; and by Dremio, which powers its Intelligent Lakehouse platform using Open Source standards such as Apache Arrow, Apache Iceberg, and Apache Polaris — making Agentic BI possible.
Agentic BI: towards more autonomous data intelligence
Agentic BI (or agent-augmented Business Intelligence) represents a paradigm shift in how organizations leverage their data.
Instead of manually building reports or dashboards, users can converse in natural language with agents capable of understanding business needs, executing the necessary queries, and returning actionable insights through a guided process.
These BI Agents make it possible to talk to data, making business analysis more intuitive while reducing the alignment effort required between teams to understand data — ultimately accelerating and improving decision-making.
However, this apparent autonomy can only work with semantic consistency, unified data access, and strong governance. BI Agents depend on a semantic and logical layer (previously scattered across reports and data pipelines) that must be exploited consistently throughout the entire data lifecycle.
In the BI world, rigor is essential; thus, AI will only have value and utility if the data it manipulates is reliable, traceable, and standardized.
This is where the Open Source ecosystem plays a decisive role.
Open Source: standardizing the semantic exchange layer and aligning BI & AI
The MetricFlow engine from dbt Labs, recently released under the Apache 2.0 license, allows organizations to centralize business metric definitions within a single semantic layer.
Each metric (e.g., KPI such as revenue, number of customers, etc.) is defined once, then automatically translated into consistent and auditable SQL queries.
This openness aligns with the dynamic initiated by Snowflake and its partners through the Open Semantic Interchange (OSI) initiative — a new open standard for sharing metric definitions across platforms and tools, ensuring that a KPI calculated by a dashboard, an API, or a conversational agent follows the exact same business logic.
These advances pave the way for semantic interoperability across the entire data processing chain — a prerequisite for building trust in BI Agents that integrate internal data previously accessed mainly via BI dashboards with the help of Business/Data Analysts.
The agentic data platform: supporting the data mifecycle
In this rapidly evolving landscape, Dremio plays a key role thanks to its Open Source-based foundations, federated through its Intelligent Lakehouse platform:
- Apache Arrow : an in-memory columnar format that standardizes data exchange between systems and dramatically boosts computation and analysis performance.
- Apache Iceberg : an open transactional table format designed for data lakehouses, ensuring reliability, versioning, and governance at scale.
- Apache Polaris : a unified and open metadata catalog, simplifying data discovery, management, and traceability.
These components make Dremio far more than a query engine — it’s an Open Data Platform of the Lakehouse type, enabling connection, virtualization, and querying of all kinds of data wherever they reside, without the need for duplication. It can be deployed across Public / Private / Sovereign Clouds and On-Premises environments.
The platform provides a native semantic layer that integrates with MCP connectors for AI Agents, enabling smooth, controlled, and seamless data usage across a diverse ecosystem of tools — all aligned on emerging open standards.
This represents the evolution toward an Agentic Data Platform — the natural extension of the open data platform, built to support AI Agents and the new generation of data-driven use cases, including Agentic BI.
An open ecosystem highlighting Smile’s integration expertise
Opening up code and standards is not just a technical choice — it’s a strategic one.
In a world where algorithms increasingly depend on data shared across multiple stakeholders, Open Source stands as the guarantor of transparency, trust, and resilience.
Integrating these components into a unified ecosystem opens new possibilities for organizations seeking to industrialize Agentic BI while maintaining control over their data.
- dbt / MetricFlow provides semantic clarity: metrics are documented, shared, and interpreted uniformly through open standards like OSI, ensuring integration into a broader ecosystem
Dremio, through Arrow, Iceberg, and Polaris, delivers flexibility and sovereignty: data remains in open formats, queryable via MCP, and integrable into intelligent agents.
This allows enterprises and organizations to build truly hybrid data intelligence: BI agents converse with the semantic layer, while Dremio bridges that semantics with distributed, organized, and governed data sources.
Agents can query data without relying on proprietary models or losing traceability — enabling explainable, auditable, and high-performing AI across use cases once limited to traditional BI.
This model fosters innovation, reduces vendor lock-in, and enhances synergy: organizations retain ownership of their data assets while accelerating the adoption of augmented BI and AI solutions.
By relying on proven technologies, interoperable standards, and the strength of the Open Source community, companies can now combine innovation, trust, and digital autonomy.
By relying on proven technologies, interoperable standards, and the strength of the Open Source community, companies can now combine innovation, trust, and digital autonomy.
This approach is deeply aligned with Smile’s values: opening rather than isolating, standardizing rather than locking in — always to innovate.
Get in touch with our experts to go further together in your Data/AI transformation.
