Effective as of 13 August 2020

Adobe Experience Platform Intelligence is an open and extensible platform that helps customers analyze and enhance data through data science and derive insights with advances queries. Adobe Experience Platform Intelligence is a package that offers a subset of the functionalities included in Adobe Experience Platform.

Customer’s use of Adobe Experience Platform Intelligence is subject to the following:

Scoping Parameters.

The following Scoping Parameters apply to Customer’s Adobe Experience Platform Intelligence production purposes, without Add-ons:

• Data Lake Storage.  Customer may store up to the Total Authorized Data Lake Storage (up to 5.5 terabytes without Add-ons) in Adobe Experience Platform Intelligence.
• Data Science Workspace Users.  Adobe Experience Platform Intelligence supports approximately 5 concurrent Data Science Workspace Users, subject to the Performance Guardrails and Static Limits detailed below.
• Query Service Users.  Adobe Experience Platform Intelligence supports approximately 5 concurrent Query Service Users, subject to the Performance Guardrails and Static limits detailed below.
• Data Scanned or Written.  Customer may Scan/Write up to 250 terabytes of data per year.
• Second-generation Data Activation.  Customer may export, through batch, to any non-Adobe solution (directly or indirectly) up to 1 terabyte of Second-generation Data per year through the Data Access API (or any other means within Adobe Experience Platform Intelligence).

Performance Guardrails.

The following Performance Guardrails apply to Customer’s Adobe Experience Platform Intelligence production purposes, without Add-ons:

• Concurrent Ad-hoc Queries.  Adobe Experience Platform Intelligence supports up to 1 concurrent Ad-hoc Query across all Users.
• Concurrent Notebooks.  Adobe Experience Platform Intelligence supports up to 5 concurrent Notebooks.
• Notebooks.  Adobe Experience Platform Intelligence provides Users accessing a Notebook with up to:
  • 32 CPU Cores and 256 gigabytes of RAM of total CPU capacity across all Notebooks;
  • individual Notebooks can be configured with up to 16 CPU Cores and 112 gigabytes of RAM (taken from the cumulative CPU and RAM capacity across all Notebooks, as described above).
• Sandboxes.  Adobe Experience Platform Intelligence includes up to 5 Sandboxes.  

Static Limits.  

• Ad-hoc Queries.  Ad-hoc Queries will time out after 10 minutes.
• Data Ingested per Year.  The total quantity of data ingested into Adobe Experience Platform Intelligence per year may not exceed 3 times the Total Authorized Data Lake Storage.
• First-generation Data Activation.  Customer may export, through batch, to any non-Adobe solution (directly or indirectly) up to 5 terabytes of First-generation Data per Profile per year through the Data Access API (or any other means within Adobe Experience Platform Intelligence).
• Model Runs.  Model Training Runs and Model Scoring Runs can run for up to 24 hours.
• Scheduled Batch Queries.  Scheduled Batch Queries will time out after 24 hours.
• Query Service.  Customer may only use Query Service to: run SQL queries for the purpose of data analysis; run SQL queries to create roll-up metrics to surface directly into a BI tool; and quickly inspect data in Adobe Experience Platform Intelligence.
• Unlicensed Features.  Customer is not licensed to, and must not, access or use the Profile Service, Identity Service or Segmentation Service components in components in Adobe Experience Platform, unless separately licensed.

Definitions

Additional Ad-hoc Query Users means an add-on to Adobe Experience Platform Intelligence increasing Customer’s authorized number of Query Service Users from 5 to 10.  Should Customer license more than one Additional Ad-hoc Query Users pack, the increased capacity identified above will be added again to Customer’s overall entitlements (e.g., a customer licensing 2 Additional Ad-hoc Query User packs will be entitled to a total of 15 Query Service Users).

Additional Batch API Activation means an increase in the cumulative quantity of Second-generation Data that Customer us authorized to export, through batch, to any non-Adobe solution (directly or indirectly)), as detailed in the Sales Order.

Ad-hoc Query means an ad-hoc query generated from the Query Service Editor, PSQL Client, Customer’s BI Tool or Data Science Workspace, as applicable.

Advanced Intelligence Pack means an add-on to Adobe Experience Platform Intelligence increasing Customer’s Data Science Workspace capabilities (inclusive of capabilities included with the base Adobe Experience Platform Intelligence offering):

• Data Science Workspace Users from 5 to 10 and enabling Data Science Workspace Users to access recipes, models and services;
• number of concurrent Notebooks from 5 to 10;
• total CPU capacity across all Notebooks from 32 CPU Cores and 256 gigabytes of RAM to 128 CPU Cores and 512 gigabytes of RAM;
• addition of 12 CPU Cores, 224 gigabytes of RAM, 2 GPU Cores and 12 gigabytes of RAM of GPU capacity across all Notebooks;
• maximum amount of processing power and storage per Notebook from 16 CPU Cores and 112 gigabytes of RAM to 32 CPU Cores, 1 GPU Core, and 112 gigabytes of RAM (taken from the cumulative CPU, GPU and RAM capacity across all Notebooks, as described above)
• addition of 128 CPU Cores and 512 gigabytes of RAM in total CPU capacity to support non-Spark jobs;
• addition of 12 CPU Cores and 224 gigabytes of RAM, 2 GPU Cores and 32 gigabytes of RAM in total GPU capacity to support non-Spark jobs;
• addition of up to 20 concurrent non-Spark jobs;
• addition of 32 CPU Cores and 256 gigabytes of RAM in total head node capacity to support Spark jobs;
• addition of 96 CPU Cores and 768 gigabytes of RAM in total worker node capacity to support Spark jobs; and
• addition of up to 20 concurrent Spark jobs.

Should Customer license more than one Advanced Intelligence Pack, the increased capacity identified above will be added again to Customer’s overall entitlements (e.g., a customer licensing 3 Advanced Intelligence Packs will be entitled to a total of 30 Data Science Workspace Users).

Data Science Workspace User means any User accessing and using Data Science Workspace or Notebooks.

First-generation Data means raw data (i.e., data that has not been enriched with insights, correlations, etc. via Adobe Experience Platform Intelligence) directly ingested into Adobe Experience Platform Intelligence from Customer’s data stores or Adobe Experience Cloud solutions.

Model Scoring Runs means the processing of Customer Data through a trained model to generate predictions.

Model Training Runs means the processing of Customer Data to train a machine learning model.

Non-production Sandbox means a Sandbox that can be used exclusively for development and testing.  Non-production Sandboxes support up to Total Data Lake Non-production Storage (measured cumulatively across all of Customer’s authorized Non-production Sandboxes).

Notebook means a Jupyter notebook or a notebook offering similar functionality.

Performance Guardrail means usage limits that relate to the scoping of Customer’s proposed use cases.  Should Customer exceed a Performance Guardrail, Customer may experience performance degradation; Adobe will not be responsible for such performance degradation.  Customers who consistently exceed a Performance Guardrail may elect to license additional capacity to avoid performance degradation.

Query Service User means any User making an SQL query to Query Service through the Query Service user interface, a Notebook, Experience Platform API, or external system any for data analysis or data exploration.

Sandbox means a logical separation within Customer’s instance of Adobe Experience Platform Intelligence isolating data and operations.  Sandboxes can be either environments used for production purposes or Non-production Sandboxes. Sandboxes used for production purposes share the authorized entitlements of the base Adobe Experience Platform Intelligence production system.  Non-production Sandboxes share the authorized entitlements specified with non-production use cases.  

Scheduled Batch Queries means a scheduled query generated from the Query Service Editor, PSQL Client, Customer’s BI Tool or Data Science Workspace that facilitates petabyte scale and the ability to write results as new data sets to the Data Lake.

Second-generation Data means any data other than First-generation Data.

Scan/Write means to import data into, scan data, or write data to any other part of the Adobe Experience Platform Intelligence in conjunction with the use of Query Service or Data Science Workspace (directly or indirectly).

Scoping Parameters means usage limits that relate to the scoping of Customer’s proposed use case.  Should Customer exceed a Scoping Parameter, Adobe will work with Customer to remedy the issue leading to overusage.  Should Customer persistently exceed a Scoping Parameter, Customer may be required to license additional capacity.

Static Limit means a usage limit that relates to the technical boundaries of Adobe Experience Platform Intelligence.  Customer may not modify Static Limits. 

Total Data Lake Non-production Storage means 10% of Total Data Lake Production Storage.

Total Data Lake Production Storage means 5 terabytes plus any licensed Additional Data Lake Storage. 

Total Authorized Data Lake Storage means Total Production Storage plus Total Non-production Storage.