Prophecy adds AI data prep tools & Snowflake support

Prophecy has launched new AI data preparation tools for business users and data analysts, and has extended its AI agents to Snowflake alongside existing support for Databricks.

The products are designed to let non-technical users review an AI agent's work before relying on the output of a data workflow or chart. Users can inspect changes visually instead of reviewing generated SQL or Spark code, and the interface now opens directly to an analysis view after a question is asked.

The launch targets a common problem in corporate AI projects, where spending on models and agents has not always led to broader operational use. Many projects stall because data preparation still happens manually in spreadsheets or because requests depend on a limited pool of technical staff.

Raj Bains, Chief Executive Officer, linked that gap to the returns companies have expected from AI spending.

"Companies have poured, and continue to pour, massive chunks of their budget into models and agents, yet pilots still aren't turning into value for business users," said Raj Bains, Chief Executive Officer, Prophecy.

He added that data preparation has often been overlooked in budget decisions.

"Few are budgeting for AI data prep, and that's exactly what's led to AI projects not meeting anticipated goals or delivering the expected ROI," said Bains.

Review process

A central part of the launch is a review flow that shows business users, step by step, what an AI agent has added or changed in a workflow or chart before they approve the result. This is intended to give analysts more direct oversight of automated work without requiring support from data engineering teams.

The offering also includes a single workspace for data preparation and analysis. Within that environment, users can refine workflows and create charts without switching tools, while dashboards convert plain-language questions into visual outputs and summary findings.

Its specialised data agent draws on user prompts and metadata-based knowledge graphs to handle tasks such as preparing data workflows, onboarding data into target formats and running analysis on demand.

Bains said the product changes are intended to keep users focused on results first, while technical detail remains available if needed.

"With this release, we want to keep business users focused on answers first, with the technical detail available on demand rather than upfront," said Bains.

Snowflake expansion

The launch also expands Prophecy's platform support by bringing its AI agents to Snowflake. Organisations that keep their analytics and data operations in Snowflake can now use the same self-service workflow already offered through Databricks.

The Snowflake version lets analysts work without moving data outside the existing customer environment. That may matter for businesses that want tighter control over governance, permissions and where data is processed.

Maciej Szpakowski, Chief Technology Officer, said many AI products require customers to move information into another environment before they can make practical use of it.

"Most AI tools on the market assume you'll pull data out to somewhere else to get value from it," said Maciej Szpakowski, Chief Technology Officer, Prophecy.

He said the Snowflake deployment keeps the tool within the controls already defined by the customer's data platform team.

"Running our AI data prep tools on Snowflake ensures governance and access stay exactly where the data platform team already puts them-and wants them-and that the AI agent works within those boundaries instead of around them," said Szpakowski.

Prophecy is positioning the release as part of a broader effort to make data work less dependent on specialist engineers. In many businesses, analysts still need technical teams either to prepare datasets or to validate what automated systems have done, slowing reporting and experimentation.

By adding a visual checking layer to AI-generated workflows, the company is targeting a middle ground between manual spreadsheet preparation and full trust in automatically generated code. The approach reflects a broader trend in enterprise software, as vendors try to show that AI systems can be inspected and governed rather than treated as opaque tools.

The new AI data preparation products are available in Prophecy's Professional and Enterprise editions. Its platform also supports BigQuery, in addition to Snowflake and Databricks.