Profile: Data Scientist
Data Scientists work with Squirro products to gain insights into project data.
This work is done through Squirro’s proprietary natural language processing (NLP) library, the no-code AI Studio, and other Machine Learning (ML) services.
This page provides an overview of the Data Scientist role and links to relevant parts of Squirro’s documentation.
As a data scientist, you will have the option to work with Squirro’s proprietary natural language processing (NLP) library, the no-code AI Studio, and other Machine Learning services.
Your role is more general than Squirro’s Model Creator role.
Reference: Learn more about the Model Creator role.
Although you may assume some of the model creator’s responsibilities yourself and use Squirro’s AI Studio to create models, your role is typically to experiment and derive insights into project data.
The beginning point for any Data Scientist within Squirro who wishes to create their own models is to understand the AI studio and how to use it in the context of a project.
This is done either working alongside a model creator, or assuming the role of a model creator yourself.
Reference: Learn more about the AI Studio.
Importing ML Models#
In addition to creating your own ML models within the AI Studio, you can also import models from other sources.
If you have a model that you have created in another environment, you can import it as an MLFlow package.
libNLP and the ONNXRuntime#
This is a Beta feature. Contact Squirro Support if you encounter any difficulties importing your ONNX model via libNLP.
Alternatively, you can use the ONNXRuntime service within libNLP if your model is in ONNX format.
This utilizes Squirro’s proprietary natural language processing (NLP) library, libNLP.
You can interact with your data programmatically using the SquirroClient (Python SDK).
The basis of any project is the data that you import into Squirro.
This is typically the responsibility of the Project Creator and may be done for you already, but you also may want to import additional data for yourself.
Once you’ve loaded your data, you’ll want to familiarize yourself with the Data Processing Pipeline options available to you.
Most notably, the Pipeline Editor offers a visual drag-and-drop interface for processing your data.
Reference: Learn more about the Built-In Steps that allow you to enrich, relate, and classify your data, and more.
Visualizing Your Data#
Once you’ve loaded and processed your data, Squirro offers a powerful array of options for visualizing your data.
This is done through dashboards, which are created using the visual drag-and-drop interface of the dashboard-editor.
Reference: Learn more about the available Built-In Widgets for visualizing your data.
If Squirro’s built-in options don’t meet your needs, you can create almost any type of data visualization using React by building your own custom widget.