Integrate.ai Documentation 
========================================

**Welcome to integrate.ai**

**Revolutionize Your Data Strategy: Evaluate Third-Party Data in Minutes, Not Months**

The demand for high-quality third-party data to fuel AI and advanced models is soaring, yet the traditional evaluation process is slow, cumbersome, and fraught with data-sharing risks, often taking over a year. Integrate.ai solves this fundamental challenge by pioneering a new paradigm for data evaluation: one that enables you to explore and test external data and models without sharing sensitive record-level information. Leveraging cutting-edge Federated Learning and privacy-preserving technologies, our system allows data to remain secure with its owners while you gain critical analytical insights. This dramatic shift eliminates lengthy data-sharing negotiations, slashing the time from initial engagement to a confident investment decision.

The integrate.ai platform is built around three core, privacy-enabled functions: **Discover** relevant datasets, **Explore** content and properties, and **Test** value against specific targets. 

For data science and actuarial teams seeking deep technical analysis, the *Python SDK* is the code-level native interface. It empowers data scientists with a familiar notebook environment to perform deep exploration, modelling, and backtesting. Crucially, the SDK allows for the privacy-protected joining of your internal data with provider data, enabling deep correlation analysis and robust model building across a broad range of model types—all without ever exposing a single record. With integrate.ai, you gain the speed, security, and insight required to accelerate innovation and unlock better business outcomes from external data.

.. grid:: 3

   .. grid-item-card:: Data Evaluation 
      :link: notebooks/index.html

      Learn about our approach for data evaluation.

   .. grid-item-card:: Quickstart Guide 
      :link: quickstart.html

      Get started with integrate.ai in four easy steps.

   .. grid-item-card:: Federated Model Training 
      :link: train-overview-vfl.html

      Learn about vertical and horizontal federated learning models.

   .. grid-item-card:: Exploratory Data Analysis 
      :link: analyze-overview.html

      Explore different techniques for data analysis.

   .. grid-item-card:: Data Preprocessing & Feature Engineering 
      :link: preprocessing.html

      Learn how to use the transform session and other tools.
   
   .. grid-item-card:: Release Notes
      :link: release-notes.html

      Updated for each release.

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Getting Started

   quickstart
   deployment
   using-integrateai
   install-sdk
   data-requirements
   session

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Use Cases

   notebooks/index

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Exploratory Data Analysis

   eda-individual
   eda-intersect 

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Data Preprocessing & Feature Engineering

   preprocessing
   sample-weight
   pca

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: PRL

   prl-session

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: HFL Model Training

   train-overview
   hfl-train
   hfl-train-glm
   hfl-gbm
   hfl-predict-tutorial

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: VFL Model Training

   train-overview-vfl
   vfl-train
   vfl-glm
   vfl-gbm
   vfl-predict

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Linear Inference

   linear-inference

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Training Utilities

   batch
   checkpoints

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Metrics

   eval-metrics
   valuation-overview

.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Reference

   reference-overview
   strategy-library
   dp-overview
   schedulers
   lrt
   onprem-task-runner
   release-notes
   api