New BigQuery functions monitor skew between training and serving data, and monitor drift in serving data over time, all using SQL.

Support for model fine-tuning in BigQuery takes input text (the prompt) and the expected output, and teaches the model to mimic the behavior.

You can now query Delta Lake tables stored in Cloud Storage or Amazon Web Services S3 directly from BigQuery, without exporting or copying the data.

BigQuery supports multimodal embedding generation via Vertex AI models, and for structured data with PCA, Autoencoder or Matrix Factorization models.

BigQuery data canvas improves data analytics with a natural language-driven experience for data discovery, preparation, querying, and visualization.

With Gemini 1.0 Pro Vision in BigQuery via Vertex AI, you can analyze images and videos by combining them with your own text prompts.

We unified key data Google Cloud analytics capabilities under BigQuery, which becomes a single, unified, AI-ready data analytics platform.

Duet AI in BigQuery is now Gemini in BigQuery and includes new AI-powered assistive features to enhance productivity and optimize costs.

Differential privacy enforcement with privacy budgeting in BigQuery data clean rooms prevents data from being reidentified when it’s shared.

Now GA, BigQuery data clean rooms has a new data contributor and subscriber experience, join restrictions, new analysis rules, usage metrics and more.