Airbyte

Airbyte Adds Workspaces and HubSpot Write Capability

Airbyte announced major updates to its Airbyte Agents platform, introducing workspace functionality for per-team governance and expanded HubSpot write capabilities that let AI agents create and update contacts, companies, deals, and tickets. Workspaces provide isolated instances with workspace-specific indexed context stores, configurable access to data connectors, and administrative visibility across the organization. The Airbyte Agents Team plan (for multi-user orgs) is priced at $299/month and includes 10,000 Agent Operations (AOs) with additional AOs billed at $0.005 each. Airbyte positions these features as part of its strategy to be a data platform for enterprise agentic AI, emphasizing governance, SSO support, and least-privilege permission scopes.

Agentic Semantic Layer for AI Agents

An agentic semantic layer is a metadata layer between AI agents and a data warehouse that defines governed metric definitions, enforces access control, and exposes programmatic query interfaces (e.g., MCP or REST/SDKs). Unlike legacy semantic layers built for BI dashboards, an agentic semantic layer is designed for programmatic consumers (LLMs, AI agents, SDKs) and requires machine-readable metric definitions, programmatic discovery/querying, structural multi-tenancy, pre-aggregation to handle high query volume, and schema-as-code with version control. It generates SQL from governed definitions (not from LLMs), scopes queries per-tenant via publishable keys, and returns auditable, consistent results. The article compares existing tools (Cube, dbt MetricFlow, Looker, AtScale, ThoughtSpot, Bonnard) and explains how the agentic layer fits into the modern data stack on top of ingestion, warehouses, and dbt transformations.

Open-source apitap moves 10M rows in 9.9s

The author published apitap, an open-source Rust-core with Python bindings engine that copies whole tables between databases with no config. Using the pip-distributed wheel on a 16-core machine, apitap moved 10 million rows from Postgres to ClickHouse in 9.9 seconds and sustained 100M rows / 46 GB in 139 seconds (1.2 TB/hour) in later runs. apitap supports Postgres→Postgres, Postgres→ClickHouse, and MySQL→ClickHouse/Postgres routes, uses bounded streaming memory to run in very small containers (0.5 vCPU / 256 MB), stages into shadow tables for atomic swaps, and includes a reproducible benchmark comparing it to dlt and ingestr. The project is MIT-licensed and open for contributors; additional connectors and platforms are on the roadmap.