ZettaQuant App

The ZettaQuant App is a Snowflake Native Application purpose-built to help financial institutions turn unstructured text into structured, actionable insights without leaving your Snowflake environment.

ZettaQuant enables end-to-end text analytics for financial institutions, from PDF ingestion and sentence extraction to relevancy filtering, classification, and ML model training, all powered by GPU-accelerated AI services running securely inside your Snowflake environment.

With ZettaQuant, your organization can rapidly build, deploy, and monitor sophisticated text analysis pipelines using your own data, enabling advanced use cases like financial stance detection, tone analysis, retail sentimental analysis, etc.

<p><br/></p>

**Key Capabilities**

- **AI-Driven Text Analytics**Perform large-scale relevancy filtering, sentiment analysis, and classification directly in Snowflake, delivering enterprise-grade performance even on massive datasets at minimal cost.
- **Automated PDF Processing**Seamlessly convert PDFs and ZIP files into structured document and sentence tables using our integrated PDF conversion engine with different modes of PDF ingestion (Simple docs, moderately complex docs, complex docs).
- **Custom Model Training & Deployment**Register, evaluate and persist your own Snowpark-optimised ML models for domain-specific tasks.
- **Streamlit-Powered UI**A built-in interface for interactive data selection, pipeline logging, and real-time monitoring.
- **GPU-Enabled Performance**Leverage compute pools with GPU acceleration for faster inference and training, directly within your Snowflake account.
- **Secure Native Execution**Your data never leaves your Snowflake account; all operations execute within the governed Snowflake runtime.

<p><br/></p>

**How to Get Access**

The ZettaQuant App offers a guided onboarding experience to ensure correct setup and optimal performance. Currently, the market version can only be accessed by users who have been onboarded to the ZettaQuant Native app by our team. In order to do so, please request access via the Snowflake marketplace or to initiate your premium access:

- **Contact us:** [e-mail]
- **Book your onboarding/demo session:** https://www.zettaquant.ai/#contact

<p><br/></p>

**Privileges and References**

The app may require the following privileges for full functionality (configured during setup):

1. CREATE COMPUTE POOL: Enables creation of GPU compute pools for ML training and inference.
2. EXECUTE TASK: Allows execution of job services on compute pools.
3. CREATE DATABASE: Enables internal schema and table creation for metadata management.
4. CREATE WAREHOUSE: Enables creation of an app managed warehouse.

原始信息汇总

ZettaQuant App 数据集概述

基本信息

• 提供商：ZettaQuant, Inc.
• 定价模式：免费试用 + Snowflake基础设施成本
• 试用期限：7天试用期

核心功能

• AI驱动的文本分析：在Snowflake中执行大规模相关性过滤、情感分析和分类
• 自动化PDF处理：将PDF和ZIP文件转换为结构化文档和句子表
• 自定义模型训练与部署：注册、评估和持久化Snowpark优化的ML模型
• Streamlit驱动的UI：内置交互式数据选择、管道日志记录和实时监控界面
• GPU加速性能：在Snowflake账户内利用GPU计算池进行更快推理和训练
• 安全本地执行：所有操作在受监管的Snowflake运行时内执行，数据永不离开账户

业务应用场景

• 市场分析：结合结构化市场数据和非结构化文本数据分析市场趋势
• 情感分析：对财报电话会议、央行通讯和新闻源应用自定义情感模型
• 定量分析：执行端到端ML管道进行主题分类、标签和定量洞察提取
• 经济影响分析：在Snowflake内分析经济政策、大规模报告和宏观经济文档
• 真实世界数据(RWD)：通过可定制预处理管道将大量PDF或ZIP文件转换为结构化Snowflake表
• 基本面分析：使用结构化文本管道分析公司基本面、财报和公司披露信息

技术要求

所需权限

• CREATE COMPUTE POOL：创建GPU计算池用于ML训练和推理
• EXECUTE TASK：允许在计算池上执行作业服务
• CREATE DATABASE：为元数据管理启用内部模式和表创建
• CREATE WAREHOUSE：启用应用管理的仓库创建

访问方式

• 当前市场版本仅限经ZettaQuant团队引导上线的用户访问
• 联系邮箱：[e-mail]
• 预约演示会议：https://www.zettaquant.ai/#contact

安全特性

• 通过Snowflake安全审查
• 基于角色的访问控制
• 敏感数据保护
• 事件共享审核机制

云区域可用性

AWS

• 加拿大（中部）
• 美国东部（弗吉尼亚北部）
• 美国东部（俄亥俄）
• 美国西部（俄勒冈）
• 其他8个区域