Prophecy

Enterprise AI · Data Engineering · Productivity Software

Launch

Prophecy launches DataGPT to run data workflows inside ChatGPT, Claude, Copilot

Chat answers vanish when the session ends; this keeps the underlying logic as reusable, auditable code your team owns.

• Prophecy launched DataGPT, letting business analysts run full data workflows from inside ChatGPT, Claude and Microsoft Copilot, or via Prophecy's own web app.
• Prophecy is a data preparation startup building an AI platform for business data work; founder and CEO Raj Bains says coders got AI that reaches production while business data users only got question-answering.
• Every DataGPT answer comes with a visual workflow map backed by open Git code, so analysts who can't read Python can still inspect, fix and approve each step before it runs.
• Approved workflows ship as governed, repeatable code that reruns on schedule directly on Snowflake, Databricks or BigQuery, without calling the AI model again and without added AI cost.
• A knowledge graph built from each customer's approved business rules and data catalogs lets future requests reuse definitions and exceptions instead of relearning them every session.
• The launch targets a common workaround where analysts upload spreadsheets to chatbots without IT sign-off, getting a script-generated answer that disappears, with its logic unverifiable, once the chat closes.
• Prophecy cites a Gartner forecast that half of all organizations will adopt zero-trust data governance by 2028 as unverified AI-generated data proliferates.
• By making AI-generated data workflows auditable and reusable rather than disposable, Prophecy is betting that enterprise trust, not just answer quality, becomes the deciding factor in AI adoption for data teams.