Volume 343, Issue 2 · 6 articles
Databricks Document Intelligence adds real-time REST APIs so apps and agents can parse and act on documents as they arrive.
Document workflows usually rely on asynchronous batch pipelines, delaying decisions in interactive or human-in-the-loop settings, so Databricks now offers Document Intelligence real-time REST APIs.
Applications can call ai_parse_document, ai_decide, ai_classify and ai_extract when a file is uploaded to understand it, extract data and route it.
The APIs can be embedded in Databricks Apps, custom agents or workflow automations.
Databricks launches Predictive AI Functions for forecasting, anomaly detection and driver analysis without ML expertise.
Databricks introduced Predictive AI Functions that let business users and SQL analysts create forecasts, detect anomalies, analyze metric drivers and make predictions without advanced ML skills or building and maintaining models.
The launch includes an upgraded ai_forecast plus new ai_top_drivers, ai_detect_anomalies, ai_predict_value and ai_predict_class.
The functions work in Genie and can be embedded in existing workflows.
Atlassian cut ingestion costs 80% and sped queries up to 95% by standardizing security telemetry on OCSF in Databricks.
Atlassian's security pipeline absorbed 44 billion events a day in many logging formats inherited from years of acquisitions.
By standardizing telemetry on the Open Cybersecurity Schema Framework (OCSF) in a Databricks security lakehouse, it cut ingestion costs by 80%, and more than 500 production threat-detection signals now run on a single schema.
Atlassian reports a 60% increase in analyst productivity and queries up to 95% faster.
Databricks makes multi-statement transactions GA across Delta Lake and Apache Iceberg tables.
Databricks announced general availability of multi-statement transactions, letting organizations run complex multi-step workflows with atomic guarantees.
Many warehouse workloads need changes across multiple tables to succeed or fail together, which historically required separate warehouse systems that added data movement, duplicate copies and fragmented governance.
Databricks says it is the first lakehouse to support multi-statement transactions natively across both Delta Lake and Apache Iceberg tables.
Stripe Data Pipeline now streams real-time Stripe data into Databricks Lakebase Postgres with no code.
Stripe Data Pipeline now supports real-time sync of Stripe data directly into Databricks Lakebase Postgres without code, removing the need to build and maintain custom API or webhook integrations.
Applications can gate features, show live billing details and trigger personalized offers with simple SQL against current Stripe data.
Databricks pitches the feature for businesses whose apps must react to payment and subscription activity at runtime.
Databricks shares benchmarks and sizing guidance for Lakebase Search hybrid vector and keyword search in Postgres.
Lakebase Search runs vector and BM25 keyword search as Postgres extensions inside the application database, avoiding a separate search system, and Databricks benchmarked it using an internal app and up to 4 million passages.
Sizing the local file cache to match the vector index kept 95-98% of pages in memory and cut warm vector latency to single-digit milliseconds.
A 14-minute load test serving 109,000 requests used just $0.02 of compute, and Databricks recommends a 0.5 CU always-on floor for production.