ClickHouse Cloud Executable UDFs GA: Native Code, Network Access, and Query-Level MetricsLLM observability has a data problem hiding inside a measurement problem. Teams collect prompts, completions, model names, latency, and trace identifiers, but token-usage fields are not always present4 min read·aidatalegends.hashnode.dev1d ago
How IFCO Cut a 7-Hour dbt Job to 2 Hours 20 Minutes on DatabricksA data pipeline does not need a radical rewrite to become dramatically faster. Sometimes, the biggest gains come from making an incremental model truly incremental—and finding out what the query engin4 min read·aidatalegends.hashnode.dev1S1d ago
Fivetran Managed Data Lake Write Credentials: Letting Engines Write to Iceberg and Delta SafelyFivetran says Managed Data Lake Write Credentials are now generally available, letting query engines run DDL and DML against lake tables. Engines authenticate via OAuth2 client credentials—one Client3 min read·aidatalegends.hashnode.dev1d ago
Netflix Conductor 4.0: Scaling Workflows to 30,000 Tasks with Asynchronous EvaluationHow Netflix rebuilt its workflow engine to go from about 2,500 to 30,000 tasks per workflow, and what it means for your orchestrator4 min read·aidatalegends.hashnode.devOct 28
OpenAI Dots: A Persistent Cloud Agent for 4,000+ AppsAn always-on cloud agent on GPT-6 Astra with 4,000+ app integrations, and why the cheaper GPT-6.1 Sol changes its economics3 min read·aidatalegends.hashnode.devOct 26