June 2026 Summaries

In the landscape of data integration, AI tools have significantly accelerated the coding process, yet they fall short of addressing the comprehensive needs for production-ready data connectors. While AI can quickly generate the basic code necessary to connect data from APIs or other sources to a data lakehouse, it lacks the capability to manage ongoing tasks such as authentication, pagination, and schema changes. Companies like Fivetran offer a solution by providing a robust infrastructure and a suite of tools, including native connectors and the AI Connector Agent, which automate much of the process while ensuring reliability and maintenance support. For those with more complex requirements, Fivetran's Connector SDK allows for custom solutions with AI-assisted development, although it still demands a commitment to maintain the custom logic. Despite AI's advancements in speeding up the initial stages of connector development, the underlying infrastructure and maintenance challenges remain, highlighting the necessity of a platform like Fivetran that offers both flexibility and managed services to ensure seamless data integration.

Fivetran and dbt Labs have introduced the Agents Schema, an open standard designed to provide AI agents with reliable business context by leveraging an Open Data Infrastructure. This infrastructure is built on open standards that ensure data is portable, reusable, and governed across various tools and environments, eliminating the need for multiple data copies that can lead to inconsistencies. The Agents Schema acts as a shared context layer within a data warehouse or lake, offering AI agents access to consistent metric definitions, semantic models, and business logic documented in plain SQL tables. This approach allows agents to query trusted information before executing tasks, ensuring they operate with the same business logic as human analysts. Fivetran facilitates reliable data movement into open data lakes with formats like Apache Iceberg and Delta Lake, while dbt provides the semantic and transformation capabilities needed to define and maintain business metrics and logic consistently. Together, these solutions empower organizations to extend openness beyond data storage, enabling AI systems to work from a unified and accurate business context.

Open Data Infrastructure (ODI) offers a strategic framework for building a flexible, governed data architecture that can adapt to diverse workloads, particularly in the evolving artificial intelligence (AI) landscape. Unlike traditional data lakes and proprietary warehouses, ODI separates storage, file formats, table formats, compute, and trust into distinct layers, enhancing flexibility while maintaining data integrity. This separation allows organizations to store data in open formats and apply metadata for governance, enabling multiple compute engines to interact with the same data without vendor lock-in. The integration of tools like Fivetran and dbt facilitates reliable data movement and governance, ensuring that data remains consistent, accessible, and ready for AI applications. By focusing on open standards and modular infrastructure, ODI helps organizations avoid the risks of inflexible architectures, positioning them to better adapt to future AI demands without compromising data trustworthiness.

Fivetran has been recognized as a Leader for the fifth consecutive year in Snowflake's 2026 Modern Marketing Data Stack report, underscoring the critical role of automated data movement in unifying data and enhancing decision-making for AI readiness. This accolade highlights the importance of trusted, governed data infrastructures as organizations transition from AI experimentation to production, requiring consistent access to reliable data. Fivetran, alongside dbt Labs, contributes to building Open Data Infrastructure, offering organizations ownership over their data and AI stack, and in combination with Snowflake's AI Data Cloud, supports automation and business logic consistency across enterprise data. The collaboration between Fivetran, dbt, and Snowflake enables organizations like Snowflake, Zendesk, and Canva to eliminate data silos, accelerate insights, and establish open data foundations that support both current analytics and future AI innovations, demonstrating real-world applications of these technologies in creating unified and flexible data environments for enhanced analytics and decision-making.

Fivetran enhances enterprise file ingestion by addressing the operational complexities of managing diverse and unstructured data formats beyond typical schema and type management. Its capabilities include filtering, which allows for declarative rules to exclude unnecessary data from reaching its destination, and hybrid deployment, which supports file ingestion from both cloud and on-premises sources using a local read agent. The platform handles multiple structured file formats such as CSV, JSON Lines, Parquet, Avro, ORC, and XML, reducing the need for upstream format conversion. Fivetran also accommodates unstructured data, enabling integration of documents and text alongside structured data with metadata for lineage and control. Moreover, Fivetran's extensibility through its Connector SDK allows for the ingestion of non-standard formats like EDI, HL7, and fixed-width records, ensuring that even legacy data systems are supported within its managed infrastructure.

Fivetran's file connectors simplify the traditionally complex process of file ingestion in enterprise data systems by providing an intelligence layer that automates tasks such as schema inference, metadata enrichment, primary key definition, type detection, and schema evolution. This approach eliminates the need for manual schema definitions by automatically inferring file structures and configuring destination tables, thereby reducing the engineering workload required to manage multiple file sources. Fivetran enhances data integrity and processing efficiency by adding metadata columns, allowing user-defined primary keys to manage duplicates, detecting data types automatically, and accommodating schema changes without manual intervention. These capabilities collectively transform file ingestion from a labor-intensive activity into an automated process, ensuring data consistency and reliability across various file formats and sources.

Charles Wang's guide outlines a detailed approach to building an AI agent by leveraging foundation models connected to secure business data and equipping them with a focused set of tools for specific, high-value workflows. It emphasizes the importance of starting with a solid data foundation that supports real-time updates and scalable ingestion into modern data lakes or lakehouses using open table formats. The process involves choosing a targeted use case, centralizing and modeling data, defining the context, exposing interfaces, and validating the agent's performance through testing and feedback loops. The aim is to automate specific bottlenecks with agents that perform reliable tasks within controlled environments, progressing from simple information retrieval to more complex, bounded actions, always ensuring human oversight in high-risk scenarios. The guide stresses the need for iterative improvements and the creation of reusable patterns to expand the agent's capabilities effectively and safely.

Open Data Infrastructure (ODI) is an architectural approach that emphasizes the use of open standards to enhance security and control over data, compute, cost, and shared context. Traditionally, security teams have been wary of open systems due to concerns about exposure, but ODI argues that open standards provide more control by allowing consistent application of access controls, lineage, and policy enforcement across the data foundation. As organizations increasingly adopt AI and other tools, the need for a flexible yet secure data infrastructure grows, and ODI meets this demand by separating storage and compute, reusing business definitions and semantics, and ensuring reliable and governed data access. With open standards, security controls are integrated into the infrastructure layer, enabling better auditability and trust, as security teams gain full visibility into data flow, and autonomous systems can access data safely with enterprise-grade controls. By maintaining control as systems evolve, ODI allows organizations to adapt without sacrificing governance, offering a secure, flexible path forward in a rapidly changing technological landscape.

Open Data Infrastructure (ODI) is an architectural framework that emphasizes interoperability by storing data in open formats within a data lake, enabling its use across various tools, compute engines, and AI systems without vendor lock-in. This approach allows organizations to maintain flexibility and adaptability in rapidly evolving markets, offering what is described as freedom of action, where individual components of data architecture can be altered independently. While managed services provide convenience by automating operations outside a company’s core expertise, they can also lead to dependency on single vendors due to proprietary formats and bundled services. ODI counteracts these limitations by offering technical, economic, and strategic leverage, allowing teams to adopt better tools, choose cost-effective solutions, and remain independent of vendor roadmaps. However, maintaining interoperability requires disciplined modularity and open standards to ensure meaningful choice, enabling organizations to adapt their infrastructure as needs and technologies evolve.

ANASAC, a leading agroindustrial enterprise in Latin America, undertook a significant digital transformation by consolidating its fragmented ERP systems across 19 countries onto SAP S/4HANA Cloud via RISE. This move aimed to unify operations and enhance decision-making capabilities by providing a comprehensive view of the entire business. To enable more effective data analytics and operational use, ANASAC partnered with Evolve Decision Science and Fivetran to extract and transfer critical SAP data into Snowflake, overcoming common data silos and technical barriers. This strategy allowed ANASAC to maintain control over their data environment during the ERP transformation, giving the company a robust foundation for future analytics and AI initiatives. Their successful implementation highlights a broader trend of Latin American enterprises investing in ERP modernization to manage complex international operations, offering valuable insights for other multinational companies considering similar transitions.

Fivetran's approach to minimizing customer churn leverages AI by centralizing and synthesizing diverse customer data sources such as Gong, Salesforce, Zendesk, and product usage data into a unified platform. This centralization, facilitated by Fivetran, allows for a comprehensive view of customer interactions and behaviors, enabling proactive retention strategies rather than reactive churn analysis. By integrating this data into a single framework, AI can more effectively analyze patterns and provide actionable insights, shifting the focus from post-churn reviews to real-time intervention opportunities. The success of this system hinges not on the AI model itself but on the combination of holistic data and embedded human expertise, allowing AI to scale expert judgment and improve retention outcomes.

Fivetran, a data integration platform, is now available on Google Cloud in the Dammam region of Saudi Arabia, allowing organizations to keep data in-region while complying with the Kingdom's Personal Data Protection Law (PDPL). This expansion removes the data residency barrier, enabling enterprises to securely and efficiently centralize data from various sources using over 700 prebuilt connectors without sending sensitive information outside the country. The platform's open data architecture supports automated change data capture, ensuring continuous data synchronization with reduced latency and operational overhead. Fivetran's availability on Google Cloud in Saudi Arabia helps organizations meet local regulations, accelerate AI adoption, and adopt an Open Data Infrastructure approach, all while maintaining global best practices for data security and compliance. This development allows CIOs, CTOs, and data leaders to modernize their data infrastructure confidently, aligning with both local and international standards.

Fivetran and dbt Labs have announced a collaboration to advance open, scalable, AI-ready data infrastructure through joint product innovations. These developments include the release of dbt Core v2.0, which offers enhanced performance and scalability with its Rust-based engine, along with dbt State, which reduces warehouse compute costs by identifying necessary model builds. The introduction of dbt Wizard, a tool tailored for the analytics engineering lifecycle, improves agent-assisted workflows by leveraging contextual project understanding to reduce production incidents. Fivetran's AI Connector Agent, which generates managed connectors from API documentation, and Agents Schema, a standard context layer for AI, further enhance interoperability and developer productivity. These initiatives reflect a commitment to creating an open and efficient foundation for AI-native data work, fostering innovation and collaboration within the data community.

Fivetran, dbt, and Google Cloud collaborate to create a robust data foundation that empowers AI agents to deliver reliable business insights without the need for static dashboards or manual data queries. This approach allows business users to interact with data through natural language queries, receiving actionable insights quickly. The integration of Fivetran automates data synchronization across systems, while dbt standardizes this data into consistent models, ensuring AI agents can effectively reason and provide answers. Google Cloud adds governance, storage, and query capabilities, as well as a natural-language interface, to enhance user interaction. The system captures data lineage to ensure transparency and trust, enabling AI agents to not only provide answers but also trace the origins of their data. This comprehensive setup transforms AI agents into reliable tools for analytics, operations, and decision-making, reducing manual data processing and enhancing the speed and reliability of insights.