dataaaaa!

The Paris Open Lakehouse and AI Meetup focuses on open source data engineering and lakehouse architectures. Technical sessions cover Unity Catalog, composable lakehouse building blocks for hybrid deployments, and Omnigent, a meta-harness for AI agents that integrates MLflow and OpenTelemetry traces.

dbt Core v1.12 introduces technical upgrades to prepare pipelines for v2.0. Key features include an opt-in Rust-based parser, the on_error model config for upstream failure handling, a dedicated vars.yml file, expanded Python and JavaScript UDF support, and a simplified Iceberg catalog spec.

The Paris ClickHouse Meetup focuses on real-time analytics at scale. Technical sessions highlight Equativ's ClickHouse Cloud migration journey, stream-enrichment pipeline optimization to eliminate redundant SQL workflows, and building self-improving agents with Langfuse for real-time systems.

AI Analytics Week covers semantic layers for end-to-end data visibility, self-hosted architectures with zero data movement, and dashboards-as-code managed via CLI and YAML. Sessions also explore agentic development workflows, Model Context Protocol integration, and governance controls.

Anthropic's data engineering team uses Claude Code workflows to automate incident response and triage operational signals. Engineers construct Claude-powered failure investigation systems, class-wide remediation tools, and automated PR review bots to catch pipeline errors and enforce requirements.

A live demonstration shows building an end-to-end data pipeline using a single unified schema. An AI agent uses dltHub to generate and deploy ingestion pipelines from a REST API into MotherDuck serverless warehouse.

BI tools face high costs and declining value as executives question whether AI prompts can replace traditional reporting. Ryan Dolley and Jacob Matson analyze the modern BI landscape and the reality of AI-driven analytics.

AI models generate data pipelines quickly, but non-deterministic code and unseen data lead to wrong schemas, units, and silent bugs. Successful data engineering requires specific habits to validate generated logic.

Data Outpost is a builder-led conference focused on the data layer behind AI systems. Key technical themes cover agentic data infrastructure, building context layers for trust and provenance, and evolving data engineering skills for agentic workflows. "Simplicity wins" highlights the shift.

Data engineering sessions at the event focus on scaling trusted self-service, building organizational context layers for AI agents, and governing semantic layers.

DuckCon in Amsterdam focuses on the technical evolution of DuckDB, DuckLake, and Quack. Engineers from organizations like Snowflake and Spotify discuss open-source analytics and data engineering patterns. The event highlights how to learn the latest developments about DuckDB, DuckLake and Quack.

This workshop demonstrates building a reporting layer for AI agents using dltHub Pro to ingest and normalize agent traces. Engineers learn to capture tool calls and token usage into structured schemas.

Data engineering shifts toward open standards like ODCS for data contracts and ODPS for data products to enable automation. By defining output ports and semantic layers in standard-conformant files, engineers can use coding agents to build architectural units. A data contract exists to build trust.

Flights enables data engineers to build and schedule Python pipelines using AI agents. It supports ingestion via MCP servers, SQL table functions, and a dedicated UI. The tool uses Flight Plan templates for production deployment and integrates with Dives to transform raw data into dashboards.

Experts analyze modern data architectures by comparing theoretical models with field realities. Key technical focuses include the maturity of lakehouses, the adoption of Apache Iceberg and open table formats, and the balance between batch and streaming.

Data engineering remains the foundation of effective AI development. Research involving hundreds of controlled experiments shows that well-modeled schemas and clear naming conventions outperform complex RAG pipelines and agent architectures.

The DiveMaxxing contest focuses on interactive data visualizations built using AI agents integrated with MotherDuck. Engineers use Claude or ChatGPT to perform live queries against datasets, moving beyond static charts. One example includes a star map built on ESA Gaia data.

The MotherDuck MCP server enables AI agents to execute SQL queries and build data artifacts directly through conversational interfaces. This shift toward agentic analytics allows for automated data exploration while maintaining human oversight.

Context engineering for data agents is most effective when focusing on well-modeled schemas rather than complex RAG pipelines. Experiments show that clear naming and simple architectures outperform expensive retrieval systems.

The ClickHouse France User Group focuses on high-performance analytical database management. Technical sessions explore real-time data processing and architectural patterns for large-scale telemetry. Attendees discuss optimizing query performance and managing distributed clusters.

Modern data engineering leverages AI copilots to accelerate the creation of robust pipelines using dltHub. Engineers can automate ingestion from sources like GitHub while maintaining control over production readiness. Elvis Kahoro demonstrates that speed matters, but so does getting it right.

The ClickHouse Paris Meetup focuses on high-performance analytical database engineering. Experts from Equativ and ClickHouse discuss real-world implementations and technical updates. Attendees dive into what is new in the world of ClickHouse to optimize large-scale data processing and storage.

Fivetran and dbt Labs are merging to build open data infrastructure focused on the agentic era. This integration aims to streamline data ingestion and transformation workflows.

Modern data engineering focuses on semantic layers to provide context for AI agents, ensuring reliable self-service by resolving metric ambiguity. Architects must also balance frugal design with AI integration when modeling large datasets, such as filtering millions of entities into qualified

Snowflake adaptive warehouses automate resource scaling to optimize performance and cost. Engineers must understand the specific billing mechanics and configuration nuances to avoid unexpected charges.