ZQ Intelligence is an orchestrator-specialist agent system for financial analysis built natively on Snowflake. It uses ZettaQuant's Model-as-a-Service (MaaS) to deploy 50+ domain-specific NLP models for tasks like stance detection, uncertainty classification, and forward-looking language analysis. A Cortex Agent orchestrator routes natural-language queries to specialist models deployed on Snowpark Container Services, then synthesizes evidence-grounded results. Two production agents are covered: ZQ Macro Agent for central bank communications and ZQ Equity Agent for earnings call transcripts. Every analytical claim is traceable via deterministic identifiers, persisted evidence tables, structured tool outputs, and execution logs — all within Snowflake's governed environment without data movement.
Table of contents
ZQ Model as a Service: How It Powers the Snowflake IntelligenceWhy Specialist Agents: Assembly Line LogicArchitecture at a GlanceUse Case 1: ZQ Macro AgentSort: