• •Spice.ai founders Luke and Phillip announced version 2.0 of their open-source engine on Hacker News.
• •The tool leverages Apache DataFusion and Ballista to provide SQL, hybrid-search, and LLM-inference across disaggregated storage.
• •Enterprise users including Barracuda Networks and Twilio are currently utilizing the engine for data infrastructure.
• •The extent to which v2.0 maintains performance parity during complex, high-concurrency analytical queries remains to be verified by third-party benchmarks.

Spice.ai has launched v2.0 of its open-source data and AI engine, designed to facilitate real-time analytical queries on operational data without requiring traditional ETL processes. Building on the foundation of Apache DataFusion, this iteration expands the tool’s ability to handle hybrid search and LLM inference across distributed storage environments. Unlike many conventional data pipelines that rely on batch-processing for analytics, Spice.ai targets direct interaction with production data. However, the operational overhead of managing these real-time queries at significant scale is not yet fully documented, leaving potential users to weigh integration simplicity against long-term maintenance requirements.