Introducing PhoenixAI Built for Agents That Query, Reason, and Act
One user prompt can turn into dozens of SQL queries — none of them reviewed before the agent acts. → See PhoenixAI (managed StarRocks): https://bit.ly/4ycMopx → Questions about your workload? 👉 https://bit.ly/4zcAhJH That's the workload AI agents put on your data layer, and it's not what warehouses, real-time OLAP, or query engines were designed for. In this video, Sida Shen (PM at PhoenixAI, formerly CelerData) walks through where each part of today's stack runs out of room and what we built instead: one analytical database that returns sub-second results on normalized tables, no precomputation required. You'll also see how Conductor, Demandbase, and Coinbase run it in production, plus a first look at our roadmap for the full agent loop. -------------------------------------------------------------------------------------------------------------------- Timestamps 00:00 Intro 00:36 CelerData Is Now PhoenixAI: Behind the Name Change 01:32 Agenda 02:26 The Shift: Why AI Agents Change What the Analytical Data Layer Must Support 06:01 The Four Core Requirements for an AI-Ready Analytical Data Layer 07:46 Where Today’s Stack Runs Out: Where OLAP, Warehouses, and Query Engines Hit Their Limits 12:14 When Query-Time Joins Fail, Precompute Becomes the Workaround — and Denormalization Becomes the Cost 14:24 What We Built: One Analytical Database for AI Applications and Agents 18:41 Query Planning: How Agent SQL Stays Fast 20:57 Streaming Upserts: Fresh and Fast at Once with Primary-Key Tables 23:08 Governance: Governed Enterprise Data Access 24:07 Side by Side: PhoenixAI vs. Real-Time OLAP, Cloud Warehouses, and Query Engines 24:57 PhoenixAI Architecture Deep Dive: One Engine for Real-Time Data and Your Lakehouse 25:43 PhoenixAI in Production: Who Runs This Today? 25:54 Conductor: The LLM Reasons Over Language; PhoenixAI Handles the Numbers 27:38 Demandbase: Customer-Facing Analytics and Agents, Powered by One Engine 29:09 More Customer Use Cases: Yuno, Celonis, and Coinbase 30:35 PhoenixAI Roadmap: Connecting AI to the Analytical Data Layer 32:09 Five Capabilities, One Database: Semantic SQL, Context Base, Agent Observability, and Agent Governance --------------------------------------------------------------------------------------------------------------------- Learn more at https://www.phoenixdata.ai/ Connect with us: LinkedIn: https://www.linkedin.com/company/phoenixai-data/ Twitter: https://x.com/phoenixdataai CelerData Website: https://www.phoenixdata.ai/ StarRocks GitHub: https://github.com/StarRocks/StarRocks StarRocks Website: https://www.starrocks.io/ Slack: https://starrocks.io/redirecting-to-slack #AgenticAnalytics #DataArchitecture #DataEngineering #AIAnalytics #AgenticAnalytics #LLM #OpenAI #ChatGPT #Claude #AIAgents #AnalyticsEngineering #DataInfrastructure #AISearch #ConversationalAI #DataStack #MCPServer #DataAnalytics #RealTimeAnalytics #RealTimeData #OLAP #DataAnalyst #DataEngineer #DataInfrastructure #databaseprogramming