Kosha.AI

In today's data-driven landscape, the ability to efficiently manage, understand, and leverage your data within Snowflake is paramount. However, organizations often grapple with outdated documentation, complex governance, inefficient data onboarding, and the challenge of making data truly accessible for informed decision-making.

​​

The Challenge: Bridging the Gap Between Data and Value

​

As data ecosystems grow, so do the complexities. Your teams may be facing:

​

• Documentation Deficits: Keeping data documentation current and accurate is a constant battle, leading to knowledge silos and slowed analytics.

• Inefficient Onboarding: Integrating new data sources and solutions can be time-consuming and resource-intensive.

• Data Quality Concerns: Ensuring the reliability and trustworthiness of data is critical for confident decision-making.

• Opaque Governance: Understanding who has access to what data, and how, can be a significant hurdle for security and compliance.

• Performance Bottlenecks & Cost Overruns: Identifying inefficient queries and optimizing warehouse usage is crucial for managing Snowflake costs effectively.

• Barriers to Data Access: Business users often rely heavily on technical teams to extract insights, creating delays.

​

Introducing the Kosha.AI - Document Curator & Analyst: Your Integrated Snowflake Command centre. 

​​

Key Capabilities for Strategic Data Advantage:

​

1. AI-Powered Documentation & Onboarding:

   • Automated Documentation: Leverage Snowflake Cortex LLMs to automatically generate comprehensive documentation for your views, stored procedures, and functions. Significantly reduce manual effort and ensure up-to-date, trustworthy documentation.

   • Streamlined Onboarding: Efficiently onboard new data solutions, define their scope, and configure automated documentation processes with customizable AI prompts.

   • Editable & Versioned Docs: Maintain control by editing AI-generated documentation, with previous versions archived for traceability.

2. Intelligent Data Quality (SmartDQ):

   • Proactive DQ Monitoring: Define, manage, and monitor data quality rules directly within Snowflake. Utilize AI assistance to translate natural language requirements into actionable DQ configurations.

   • Comprehensive Dashboards: Gain at-a-glance insights into your data quality posture and trends over time.

   • AI-Enhanced View Analysis (SmartDocs for DQ): Automatically generate functional and technical descriptions, improvement suggestions, and column details for your views to aid in quality assessment.

3. Enhanced Governance & Security Visibility:

   • User & Role Governance Explorer: Gain unprecedented clarity on data access. Interactively visualize complex Snowflake role hierarchies, user assignments, and object permissions to ensure robust security and compliance.

   • Application Role Management: Centrally manage application-level personas and permissions, mapping them to Snowflake roles for controlled feature access.

4. Performance & Cost Optimization:

   • Warehouse Performance Insights: Monitor warehouse credit consumption to identify optimization opportunities.

   • AI-Driven Query Optimization: Automatically identify long-running queries and leverage Cortex AI to receive actionable optimization suggestions, reducing costs and improving performance.

5. Democratized Data Access & Semantic Layer:

   • Semantic Model Generator: Easily create and manage semantic models (YAML definitions) for Snowflake Cortex Analyst. Define tables, metrics, and relationships to provide business context to your data.

   • Chat with Your Data: Empower business users to ask natural language questions directly against your data through deployed semantic models, fostering a data-driven culture.

6. Accelerated Development:

   • Smart Code Generator: Utilize AI to generate data ingestion code (PySpark, Airflow, Python, SQL) from textual descriptions or even uploaded flow diagrams, speeding up development cycles.

​​

Why Smartello for Your Snowflake Ecosystem?

​

The Smartello Document Curator & Analyst empowers your organization to:

​

• Maximize Efficiency: Automate mundane tasks and accelerate data onboarding and development.

• Improve Data Trust: Enhance data quality and ensure documentation is current and reliable.

• Strengthen Governance: Gain clear visibility and control over data access and security.

• Optimize Snowflake Investment: Reduce warehouse costs and improve query performance.

• Empower All Users: Make data more accessible and understandable for both technical and business teams.

• Innovate Faster: Leverage cutting-edge AI capabilities directly within Snowflake.

Transform Your Data Management Today

​

Stop letting data complexity hinder your progress. The Smartello Document Curator & Analyst provides the intelligent, integrated solution you need to take full control of your Snowflake data and analytics layer.

​

Ready to elevate your Snowflake data strategy? Contact Smartello today to learn more or request a personalized demonstration.

​The application role (e.g., SMARTELLO_APP_ROLE) requires:

​

USAGE on the application package.

• SELECT on SNOWFLAKE.ACCOUNT_USAGE views (e.g., USERS, ROLES, GRANTS_TO_USERS, GRANTS_TO_ROLES, WAREHOUSE_METERING_HISTORY, QUERY_HISTORY).

• SELECT on INFORMATION_SCHEMA views across databases/schemas the app needs to access.

• USAGE on databases and schemas, and SELECT on tables/views that will be documented, analyzed for DQ, or included in semantic models.

• USAGE on the database/schema of the SEMANTIC_MODEL_STAGE and CODEGEN_TEMP_STAGE.

• WRITE (or CREATE FILE / PUT) on SEMANTIC_MODEL_STAGE and CODEGEN_TEMP_STAGE.

• SHOW SEMANTIC MODELS account-level privilege.

• USAGE on functions SNOWFLAKE.CORTEX.ANALYST,

• SNOWFLAKE.CORTEX.COMPLETE, and any custom Cortex Functions like CORTEX_OPTIMIZER_FUNCTION.

• Full DML/DDL privileges on the CFG schema and its objects.

​

Application Configuration Parameters (Python Script Constants)

​

• SEMANTIC_MODEL_STAGE: Snowflake stage path for saving semantic models (e.g.,   @MY_DB.MY_SCHEMA.MY_SEMANTIC_MODELS_STAGE). Must be updated by admin post-install.

• CORTEX_ANALYST_FUNCTION: Fixed as SNOWFLAKE.CORTEX.ANALYST.

• CORTEX_OPTIMIZER_FUNCTION: Fully qualified name of the Cortex Function used for query optimization (e.g.,

• SMARTDOCUMENTER.CFG.OPTIMIZE_QUERY_WITH_CLAUDE35_SONNET). Can be updated by admin if a custom function is preferred.

• MAX_TREE_DEPTH: Integer controlling recursion depth for governance tree views.

• INDENT_CHAR: Character(s) used for indentation in some outputs.

• cfg_schema: Name of the schema used by the application (e.g. SMARTDOCUMENTER.CFG).

​