Process Types
AA Loader
The Adobe Analytics or AA Loader process provides the ability to bring raw Adobe Analytics clickstream data into a customer environment from the specified location where Adobe places the files (i.e. S3 bucket or SFTP folder). Once the hit_data and lookup files are placed into a specified location…
Adobe API
The Adobe API process provides the ability to pull in the Adobe custom variable definitions into the Adobe Analytics Input Adapter app creating a Report Schema output that provides the details and rules for each eVar, Prop, and Event. Custom variable definitions allow the ability to create…
Adobe Audit
Adobe Audit is one of the many processes available in the Syntasa platform that users can utilize while creating apps. The process is available in all three modules, Synthesizer, Composer, and Orchestrator.
Adobe Classifications Export
The Adobe Classification Export process allows us to request a classification report from Adobe via API calls. To utilize this process, the Adobe API connection needs to be configured. Then we can make use of the connection by using the Adobe API process.
Adobe Event Enrich
Using data prepared by the AA loader process, the Event Enrich process applies functions to the data, joins lookups, and writes the data into an event-level dataset. The dataset is the foundation for building the session, product, and visitor datasets, but can also be used for analysis and…
Adobe Product Enrich
Using data prepared by the Event Enrich process, the Adobe Product Enrich process applies functions to the data, joins lookups and writes the data into a product level dataset, which can be thought of as a sub-event level.
Alert
The purpose of the Alert process is to provide a mechanism for evaluating a variable or set of variables against a threshold and provide an email or Slack message when that threshold is met. For example, if a model is running and the analysts needs to know if the grade of the model gets to a…
BQ Process
The BigQuery (BQ) Process is a Synthesizer, Composer, and Orchestrator process that builds and manages user-defined analytics datasets using pre-written and verified working code. It is designed to allow the user to query an existing BigQuery table to create a new dataset within the Syntasa…
Custom Shell
The Custom Shell process is used for running a package that contains shell script and load their output to a dataset table.
Dash
With a focus on visuals, the Dash Process enables data scientists to create dashboards from the data processed through their pipeline rather than just dataset tables. This is an additional tool that highlights data and turning this data into dashboards, adding to the capability of dashboards.
Decision Tree
The Decision Tree process is used to create a model that predicts the value of a target variable based on several input variables. This process type provides the ability to define desired settings and fields to model in an effort to provide the user with the ability to experiment faster and…
Event Loader
The purpose of the Event Loader process is to ingest non-Adobe raw files, such as marketing data or enterprise lookup data.
Evaluator
Once the learning and scoring processes have completed, the Evaluator Process is used to produce a dataset for analytical use.
Feature Learn & Feature Transform
When dealing with different data science use cases, we come across different kinds of variables such as strings, numbers, or text data. Datasets usually have a mix of categorical and continuous variables. However, algorithms that are powered by these datasets understand only numbers. Hence, it is…
Featurize
Featurize is a composer process that helps in performing feature engineering on the Input datasets. Some of the high-level capabilities available with Featurize are:
From BQ
The From BQ provides the ability to load data from a BigQuery table into a Hive structured storage location in the Syntasa environment. Sample use cases may include:
From DB
The purpose of the From DB process is to define the input database table, schema, and load into a big data environment. A database connection will need to be defined first on the Connections page and that Database Connection will need to be dragged onto the canvas and connected to the From DB…
GA Output Adapter
The GA Adaptor provides the ability to configure and import a file to the Google Analytics Data Import. Data Import provides a means to upload external data (i.e. CRM, product catalog, model scores) to GA providing metadata that helps to add context to the Analytics behavioral data.
GA Report
A process to download a customized report from Google Analytics.
Generalized Linear Regression
The Generalized Linear Regression model is a flexible generalization of ordinary linear regression that allows for response variables that have error distribution models other than a normal distribution . The GLM generalizes linear regression by allowing the linear model to be related to the…
Generic Event Enrich
The Generic Event Enrich process applies functions to the data, joins lookups and writes the data into an event level dataset. The dataset is the foundation for building the session, product, and visitor datasets, but can also be used for analysis and user-defined analytics datasets.
Gradient Boosting Tree
The aim of the Gradient Boosting Tree process is to create or "train" an ensemble of trees, given how to train a single decision tree. This technique is called boosting because the expected outcome of an ensemble is to work much better than a single estimator. This process type provides the ability…
LTM
The LTM Process is one of the steps in the process of building the training dataset using the Lookback process dataset.
Lookahead
The Lookahead process creates the label dataset which contains the outcomes to predict. A label is a defined success event that the model learns from. It is possible to have more than one label for modeling.
Lookback
The Lookback process is used to build a historical dataset for applications like the propensity scoring app. This dataset includes all the features needed to train a machine learning model by looking back a specified number of days from the processing time and creating a table with user-defined…
Logistic Regression
The Logistic Regression process type is used when the dependent variable is dichotomous (binary). This process is used to describe data and to explain the relationship between one dependent binary variable and one or more nominal, ordinal, interval or ratio-level independent variables in an effort…
ML Engine Prediction
Integrate an existing prediction model into an app workflow, effectively productionalizing the model. The model(s) will be able to be ran on schedule.
ML Engine Train
Integrate an training model into an app workflow, effectively productionalizing the model. The model(s) will be able to be ran on schedule.
Matomo Loader
The purpose of the Matomo Loader process is to get data from the Matomo website
Orchestrator Free Form
"Orchestrator Apps" refer to applications that create integrations with enterprise systems, enabling them to deliver data points. Orchestrator Free Form is an Orchestrator app that provides a blank workflow for the ability to create custom apps by using pre-built processes like To DB , Post , and…
Post
The Post is an Orchestrator process that facilitates the building of a Customer Intelligence Hub by pushing data out to non-relational databases such as Big Table and HBase. Some of the high-level capabilities available with the Post are:
Quantile Evaluator
The Quantile Evaluator on the Syntasa platform calculates cumulative performance metrics based on score percentiles.
Quantile Evaluator by Partition
The Quantile Evaluator on the Syntasa platform calculates cumulative performance metrics by score percentile and date.
Random Forest
The Random Forest process type provides the ability to do a Random Forest algorithmic model. This process type provides the ability to define desired settings and fields to model in effort to provide the user with ability to experiment faster and productionize the model so that Syntasa manages the…
Registry Identity
The Register Identity process provides the user the ability to identify and register identity pairs with the Syntasa Identity Graph. An identity graph's contents include the identity pair, first time that pair was observed and whether that pair is suspect or not. The identity pair consists of a…
Score
Once the Learning Process has completed, the Score Process provides the ability to configure and process the scoring code into a Syntasa Composer workflow.
Session Enrich
The Session Enrich process applies functions to data generated by the Event Enrich process, joins lookups, and writes the data into a session-level dataset, which can be thought of as event data aggregated at the visit level. This data is then grouped by day to allow for partitioning, which allows…
Spark Learn
The Spark KNN Learn process type provides the ability to configure and productionize custom Spark K-Nearest Neighbors Learn code into a Syntasa Composer workflow. By configuring this process it is allowing Syntasa to manage the learning on a scheduled basis. A user will find that there are two ways…
Spark Score
The Spark KNN Score process type provides the ability to configure and productionize custom Spark K-Nearest Neighbors scoring code into a Syntasa Composer workflow. By configuring this process it is allowing Syntasa to manage the KNN model scoring on a scheduled basis. A user will find that there…
Split
The Split process is used to divide your dataset into two parts: Training and Testing datasets.
To DB
This process provides the ability to write data to a downstream database.
To File (aka Publish)
To File is an Orchestrator process that facilitates pushing data out to external systems and storage locations. Some of the high-level capabilities available with the To File process are:
Transform
The Transform is a Synthesizer, Composer, and Orchestrator process that helps to build user-defined analytics datasets. Some of the high-level capabilities available with Transform are:
Unified Event Enrich
The Unified Event Enrich process takes the data in raw and enriches and transforms it to follow the SYNTASA schema as defined in the configuration mapping unique to each data source.
Unified Product Enrich
The Unified Product Enrich process uses data prepared by the Unified Event Enrich process to apply functions to the data, connect lookups, and publish the data into a product-level dataset, which can be thought of as a sub-event level.
Unified Session Enrich
The Unified Session Enrich process applies functions to data generated by the Event Enrich process, joins lookups, and writes the data into a session-level dataset, which can be thought of as event data aggregated at the visit level. This data is then grouped by day to allow for partitioning, which…
Unified Visitor Enrich
Using data prepared by the Unified Event Enrich process, the Unified Visitor Enrich process applies functions to the data, joins lookups, and writes the data into a visitor-level dataset, which can be thought of as event data aggregated at the visitor level.
Using Code Processes
This example is meant to show:
Visitor Enrich
Using data prepared by the Event Enrich process, the Visitor Enrich process applies functions to the data, joins lookups and writes the data into a visitor level dataset, which can be thought of as a event data aggregated at the visitor level. This data is then grouped by day to provide ability to…