Start a Discovery
Browse documentation

Welcome to Neuralift

What Neuralift does, who it's for, and the path from raw warehouse data to activated customer segments.

Neuralift turns the customer data already sitting in your warehouse into segments you can act on: each one discovered by Neuralift’s patented deep-learning segmentation model (US 12,321,960 B2), explained in plain language, and ready to deliver back to your data platform or activation channels.

This guide walks you through the product end to end: connecting your data, framing a use case, exploring the segments and actions that come back, and delivering labeled results to your warehouse. Along the way you’ll see which steps you drive and which steps Neuralift’s data science team handles for you.

The golden path

A typical journey through Neuralift looks like this:

  1. Connect your warehouse. An admin in your organization links Snowflake, BigQuery, Redshift, or Databricks and syncs the source tables you want to work from. See the data connections overview.
  2. Neuralift prepares your data. Working from your synced source tables, Neuralift builds a prepared dataset and profiles it into a data dictionary of KPIs, continuous, categorical, and ID columns. See how data preparation works.
  3. Define the use case together. You and Neuralift frame the business question (for example, reduce churn or grow loyalty) as a use case with a goal, a target segment count, business context, and prioritized KPIs. See creating a use case and defining the use case.
  4. Neuralift runs and publishes. Neuralift starts the segmentation run, reviews the results with you, and publishes them so your whole team can see them. See starting a run and publishing.
  5. Explore the results. Read the segment landscape like a map of your customer base, dig into each segment’s insights and persona, and review the recommended actions and their action plans. See the discovery overview and from insight to action.
  6. Export and activate. Export segments to CSV, or deliver labeled results straight back to your warehouse for campaign activation. See the exporting overview.
  7. Ask your AI assistant (Preview). Optionally, connect Claude, ChatGPT, or Copilot to your published results over MCP and ask questions in natural language. This is available by request under agreement with Neuralift. See the MCP overview and release stages.

Who does what

Neuralift is white-glove by design: Neuralift’s data science team works alongside yours inside the product. They create use cases, prepare your data, start runs, and publish results, always working with you on the definition and the review. Admins in your organization manage the workspace: data connections, members and invitations, delivery configuration, and settings. Members explore published results and export them. The full capability matrix is in Roles & permissions.

Where to go next