AI & Machine Learning

AI & Machine Learning Overview

The AI & Machine Learning object types empower users to build, deploy, and leverage intelligent systems directly within the ARPIA platform. These tools support a wide range of use cases—from generating embeddings and training models to real-time inference and custom ML workflows using powerful compute environments. With support for both GUI-based and code-based development, users can move quickly from data preparation to advanced analytics, even across GPU-backed pipelines.

Development Environments

AutoML

This object type provides access to the Machine Learning and AI tools available on the ARPIA platform. These tools support a wide variety of functions, from generating text embeddings to training custom machine learning models tailored to meet specific needs.

• AP AutoML Engine
  A GUI-based tool that allows you to train ML models on structured datasets with zero-code setup. Ideal for regression, classification, and basic time-series tasks.

• Singular-AI Text Embeddings
  Generates vector embeddings from raw text, enabling semantic search, classification, clustering, and more. Frequently used in NLP and RAG pipelines.

• AP Generative AI Workflow
  A framework for integrating LLM-based components into your workflows. This is useful for summarization, chat agents, or content generation pipelines.

AutoML GPU

This object type provides direct access to the platform's GPU processing power, enabling efficient handling of larger, more demanding machine learning tasks.

• AP AutoML GPU Engine
  Similar to the standard AutoML Engine but GPU-accelerated. Designed for large datasets or deep learning use cases where performance is crucial.

Extract

These objects automate the extraction process from MySQL-compatible databases registered as Data Sources within the ARPIA workspace. Extractions can be configured to either transfer data directly from table to table or use custom SQL queries.

• AP DataPipe Engine - MySQL
  A GUI-based pipeline builder for extracting data from MySQL sources and loading it into destination tables.

• Python 3.12 DataPipe Engine
  A code-based object for advanced users needing dynamic or conditional extraction and loading logic using Python.

High Performance Computing

These objects provide open development environments to execute complex, nested, or compute-intensive workflows, offering full programming flexibility.

• PHP 7.4 Application
  For custom logic, integrations, or lightweight server-side workflows.

• Python 3.8 Advanced ML Application
  For creating, training, or deploying ML models in a customizable Python environment.

• Python 3.8 Advanced ML & Plotly
  Similar to the ML Application object, but with built-in Plotly support for data visualization and exploratory analysis.

• R Environment
  Ideal for statisticians or teams using the R language for analytics, modeling, and reporting.

• Python 3.9 Google Cloud Speech
  A pre-configured environment for running speech-to-text tasks using Google’s Cloud Speech API.

• NodeJS16 Application
  A full Node.js runtime environment for real-time apps, APIs, or external service integrations.

Image Analyzer

This object allows for image categorization and analysis using pretrained AutoML vision models. It works by referencing image URLs stored in a repository table within the Workshop Project.

• AP AutoML Vision Predictor
  A GUI-driven object that uses AutoML Vision models to classify images based on their content. Ideal for tagging, validation, or visual categorization.

Notification Engine

This object gives you control over ARPIA’s email delivery system. You can configure notification triggers and design custom email templates via a GUI.

• AP Notification Engine
  A GUI-based interface to define when and how email alerts are sent. Requires a Mailgun API key.

Prepare & Transform Tools

These tools are used to refine, transform, and optimize your data before feeding it into machine learning models or analysis pipelines.

• AP Prepared Table
  A GUI tool that takes an existing table and converts it into a "prepared dataset" with transformation capabilities on a field-by-field level.

• AP Transform String to Binary
  Converts categorical or binary values into 0/1 representations using a GUI.

• AP Transform String to Numeric
  Maps string values to numbers, commonly used for model training or sorting.

• AP Transform Dates to Numeric
  Converts dates into useful numeric formats like day-of-week, Unix time, etc.

• AP SQL Code Execution
  Allows you to define full SQL processes via a programmable SQL block to perform advanced logic before model training.

• AP Model Render
  Runs predictions using a trained model and writes the results into tables. This works in the post-training phase.

• SingularAI Text Splitter
  Breaks long-form text into smaller chunks for summarization, embedding, or analysis—ideal for use with LLMs or semantic search.

Web-Hook Sender

This object allows the platform to send HTTP-based webhook notifications to external services or APIs.

• AP Web-Hook Sender
  A GUI tool to define outgoing payloads, endpoints, and trigger conditions for real-time communication with other systems.