AI Data Preparation for Snowflake with Claude Code

AI promises instant data insights, yet most enterprises still cannot get to answers because their data is not clean and trusted. Up to 90% of a data team's time is still wasted on the grueling work of reading, cleaning, blending, and validating raw data. Can AI finally automate this data prep bottleneck?

Two core technologies now exist:

1. Claude Code: AI and agents provide raw intelligence and code generation that, when given the right data context, can generate business logic as raw code.
2. Snowflake: Cloud data platforms provide scalable data transformation and querying with enterprise governance.

You’ve invested in Snowflake and AI, but you cannot simply hand these tools to business teams and expect immediate productivity. Accelerating data preparation requires entirely new capabilities and a fundamental rethink of the user workflow:

1. Human-in-the-loop trust: Code agents can produce code, but code alone does not work for business data users. Users need a way to understand what AI created, refine the logic, validate the results, and trust the workflow before it drives decisions. As business data users and AI agents collaborate to prepare data, teams also need a shared workspace to review, reuse, and govern that work — a need that data product proponents have long recognized.
2. Unified workflow execution: AI agents and users need a unified data preparation execution layer as the target for generating business logic. SQL data platforms provide data transformation, but data preparation workflows also need operators to bring data in through connectors, perform specialized data processing, write to targets, trigger checks and notifications, and support production workflows with scheduling, orchestration, and monitoring.
3. Reliable data intelligence: AI agents need data context, specialization for data tasks, and the tools to generate business logic with speed and reliability. They also need ways to test whether generated logic is correct — both for the schema of the data (compiles) and for processing the data ingested by the workflows (executes) 

That is the role of Prophecy.

Prophecy: AI data preparation platform

Prophecy is the AI data preparation platform for Snowflake, uniting the core capabilities needed to make AI-generated workflows reliable, trusted, and production-ready:

• Data Studio gives users a shared visual, AI-native workspace where they can describe business intent, inspect AI-generated workflows, refine the logic, validate results, and trust the output. 
• Automate provides the unified workflow execution layer for AI and users alike. It connects to data sources, runs transformations and analysis on Snowflake, and delivers outputs to downstream targets. It provides a consistent experience from development to deployment across data platforms.
• Data Intelligence turns raw AI reasoning into reliable data-preparation intelligence, so agents can generate correct, consistent, enterprise-ready workflows.

With Prophecy, business users do not start by manually building workflows from scratch. They describe business goals, and AI generates visual data workflows that users inspect, refine, validate, and trust. The workflows run on Snowflake with enterprise scale and governance, and can move into production with scheduling, orchestration, and monitoring.

Snowflake Integration

Prophecy provides native experience on Snowflake, including inheriting governance and security. Visual workflows execute step-by-step inside Snowflake, and you can see the data after each step. Data transformations get pushed into Snowflake as native SQL code, and all the processing can happen within your Snowflake environment, with data never leaving it.

Data Studio

Prophecy Data Studio gives business data users a visual, AI-native workspace for preparing and analyzing data.

Users describe their business goal, and AI generates the visual data preparation workflow for it, which users can inspect, refine, validate, and trust. Alongside the workflow, users can ask for analysis and see it generated as text, tables, and visualizations to confirm that the prepared data is producing the right business insights.

The key technology behind Data Studio is Prophecy’s compiler-backed visual workflow layer. Prophecy represents the same business logic simultaneously as code and as a visual workflow, with instant bidirectional conversion between the two.

This is what makes AI usable for business data preparation. Code agents can generate logic in the form they are best at: code. Prophecy then brings that logic into a visual form where data users can understand it, edit it, and validate it without reading pages of SQL or Python.

Data Studio is best shown through demos:

• Generation & Refinement Demo: This demo shows a user describing a business question, AI generating a visual data workflow, and the user refining the workflow step by step. 

• Data Onboarding and Validation Demo: This demo shows AI helping onboard new or messy data into a trusted structure with user validation.

Automate

Prophecy Automate is the execution layer for AI agents and Data Studio, providing a consistent runtime environment for workflows from development to production. 

It gives AI agents and users the operators they need to bring data in, transform it, analyze it, deliver outputs, and operate the workflow reliably over time.

Automate provides a batteries-included execution layer, including connectors to read and write data, transformation in Prophecy or Snowflake, production deployment, including scheduling and observability.

Data Intelligence

Prophecy Data Intelligence turns raw AI reasoning into reliable data-preparation intelligence.

Claude Code provides powerful reasoning and code generation. Prophecy specializes in that capability for enterprise data work, so agents can always generate workflows that are correct, consistent, and guaranteed to run correctly on Snowflake.

Data Intelligence combines three capabilities:

• Enterprise data context: Prophecy grounds agents in a knowledge graph built from Snowflake metadata, source databases, SaaS systems, and prior workflow computations. 
• Data-specialized generation, learning: Prophecy agents are specialized to automate data preparation tasks with speed and consistency. As users make edits to the generated output, the AI gets better for those projects.
• Validation: Prophecy checks generated logic before it reaches the users, making sure it is correct (compiles), and it runs correctly on the data (executes).

A key part of Data Intelligence is Prophecy’s restructuring compiler. It lifts generated logic into visual workflows, turning technical constructs into business-readable operations. For example, SQL window functions can be represented as concepts such as running totals, rankings, or period-over-period calculations that business users can understand and refine.

Summary

You have invested heavily in Snowflake and advanced AI models like Claude or OpenAI. Yet, the key to unlocking their true value—reliable data preparation—remains out of reach for your business teams.

Prophecy uniquely bridges this gap by combining AI powered low-code usability with high-performance, governed cloud execution that lets you stay in control. Unlike legacy tools that force risky data egress or lock you into proprietary logic, Prophecy delivers native, agentic data engineering directly inside the Snowflake ecosystem.

As the only AI-native data preparation and analysis platform built to empower business data users, Prophecy turns your natural language goals into visual, production-grade workflows that your teams can inspect, validate, and trust.

Getting Started with Prophecy

Prophecy users are starting fresh or come from an existing product such as Alteryx or SAS Enterprise Guide. 

If you’re starting fresh, you can sign up for the Professional version for $150/user/month and start preparing data directly within Snowflake. 

If you’re an enterprise, you can get started for $4000/month for 20 users with Enterprise Express, which additionally includes single-tenant SaaS, integration with your enterprise environment, and Prophecy assistance for 90 days to get your first project successful.

For teams migrating from Alteryx, Prophecy features a native importer that typically converts over 90% of your existing workflow footprint. We can analyze your current environment to provide precise effort and ROI estimates, while our Enterprise Express project provides dedicated Prophecy experts to help modernize your first workflows.