DAILY BRIEFING · SATURDAY, JUNE 13, 2026
With Databricks' Data + AI Summit opening Monday, this Saturday's fresh moves cluster around open-lakehouse interoperability — a new Linux Foundation sharing standard, Postgres-as-lakehouse, and maturing table-format governance — while agentic automation pushes into FinOps.
⇣ Jump To
⚡ QUICK TAKES
| Story | Signal |
|---|---|
| ↗ Databricks launches OpenSharing under the Linux Foundation | Sharing standardizes for the agentic era: agent skills, models, and Iceberg recipients all under one open protocol. |
| ↗ pg_lake ships natively in Snowflake Postgres | Plain Postgres becomes an Iceberg-aware lakehouse via SQL — Snowflake doubles down on Postgres + open formats. |
| ↗ Open table formats mature through governance, not just features | Iceberg, Parquet, and Arrow are codifying decision-making — project governance now matters as much as release notes. |
| ↗ Finout debuts agentic FinOps at FinOps X 2026 | Cloud-cost control over data workloads shifts from dashboards to autonomous detect-investigate-remediate agents. |
Databricks Newsroom · June 2026
OpenSharing is the Linux Foundation successor to Delta Sharing — and the first open, vendor-neutral protocol for sharing AI assets (agent skills, models, unstructured data), not just structured tables. It adds Apache Iceberg IRC client support so providers reach Iceberg-native consumers, and lets on-prem and private-cloud data connect to cloud platforms with zero copy via storage partners including MinIO, Qumulo, and Everpure. OpenAI, SAP, Stripe, Atlassian, and LSEG are already building on it — a pointed pre-Summit move to make open sharing the agentic-era default rather than a single-vendor marketplace.
✍️ Databricks · Read article →
The New Stack · June 2026
pg_lake — the ~15-extension open-source bundle (released to OSS in November) that turns vanilla PostgreSQL into an Iceberg-aware lakehouse — is now native in Snowflake Postgres, the managed service built on the Crunchy Data acquisition. Engineers read and write open table formats in object storage through plain SQL, with the extensions handling format conversion, versioning, and caching. Paired with Horizon Catalog's expanded Iceberg management, it signals Snowflake leaning harder into Postgres-plus-open-formats as a first-class architecture rather than a side bet.
✍️ The New Stack · Read article →
Apache Data Lakehouse Weekly · June 2026
The week's roundup shows open table formats maturing through governance, not just features: Iceberg shipped four C++ release candidates in seven days, Parquet contributors ran a 40-message thread on what versioning should even mean, and Arrow wrote down rules for reviewing AI-generated code. Near-term milestones include Polaris 1.6.0 branching around June 26 and Iceberg patch releases 1.11.1 and 1.10.3 heading to votes, with parquet-format 2.13.0 finally fixing float statistics. For teams standardizing on Iceberg, these project-governance signals matter as much as the release notes — they determine how predictable the format roadmaps will be.
✍️ Apache Data Lakehouse Weekly · Read article →
Finout · FinOps X 2026 · June 2026
At FinOps X 2026 (San Diego, June 8–11), Finout launched its largest expansion to date: an agentic operating model on top of its MegaBill data layer, with three specialized agents that detect cost anomalies, investigate root causes, and orchestrate remediation. It lands as the conference reframed FinOps as full "Technology Financial Management," with data-cloud platforms and AI named the most actively managed spend categories. For data teams, autonomous cost control over Snowflake, Databricks, and BigQuery workloads is moving from dashboards into agents — the same detect-investigate-act pattern reshaping observability is now arriving in cost governance.
✍️ Finout · Read article →