Snowflake Cortex AI Gateway introduces governed end-to-end agent traces
Snowflake details governed agent tracing through Cortex AI Gateway, combining model and MCP tool activity in Snowflake with role-controlled access and optional payload capture.
Data, Infrastructure & AI Platforms
NVIDIA open-sources Run workload and GPU orchestration components
NVIDIA releases the full Run:ai platform as open source, including Kubernetes GPU scheduling, sharing and resource management, with community and enterprise support offered separately.
Microsoft says Maia 200 powers production AI while Maia 300 targets 2027 rollout
Microsoft confirms Maia 200 is serving production AI workloads and outlines Maia 300's planned 2027 deployment, including multi-gigawatt scaling and projected inference-efficiency improvements.
MLflow 3.17 adds typed AI evaluation results and faster trace analysis
MLflow 3.17 introduces structured AI evaluation decisions, granular permissions and faster trace analytics, with TypeSafe dependencies and database-upgrade requirements.
Probabilistic AI Regression Testing Needs to Know Which Differences Matter
Probabilistic AI regression testing distinguishes legitimate output variation from behaviour that should block a release. See how exact checks, criteria-based evaluation, repeated sampling and critical failure boundaries can identify meaningful regressions without treating every textual difference as a defect.
Snowflake makes Postgres-to-Snowflake data mirroring generally available
Snowflake makes PostgreSQL data mirroring generally available for its managed Postgres service, supporting continuous replication, automatic schema updates and low-lag analytical views.
Inference Cost Is an Architecture Signal, Not Just a Cloud Bill
AI inference cost architecture treats rising model spend as a signal about how an application is performing its work. See how token use, cache behaviour, model-call count, routing and execution paths can expose architectural changes before they appear as reliability problems.
Databricks expands Lakeflow Designer with visual quality checks and output operators
Databricks summarises Lakeflow Designer updates covering visual data-quality checks, deduplication, flexible joins and additional output formats, while distinguishing shipped features from upcoming integrations.
AI Platform Portability Depends on Whether Policies Can Move With the Workload
AI policy portability across platforms means preserving the governance decisions that made a workload acceptable, not copying identical configuration. See how policy intent, enforcement translation and decision-level equivalence can prevent governance drift when AI workloads move between platforms.
An AI Release Can Stay Green While Task Completion Falls
AI task success production metrics measure whether a release still completes the work users came to accomplish. See how task outcomes, release identity, delayed business evidence and task-level comparisons can expose regressions that latency, error rates and tool-success metrics miss.