A Case for Business Process-Specific Foundation Models

TL;DR

This paper advocates for developing specialized foundation models tailored to business process data, addressing unique challenges like data scarcity, multi-modal data, and privacy to improve process mining, optimization, and decision-making.

Contribution

It introduces the concept of business process-specific foundation models, emphasizing their necessity and outlining the unique challenges they aim to solve.

Findings

Business process data has unique characteristics requiring specialized models.

Proposes that tailored foundation models can improve process mining and decision-making.

Highlights challenges like data scarcity and privacy in applying AI to business processes.

Abstract

The inception of large language models has helped advance state-of-the-art performance on numerous natural language tasks. This has also opened the door for the development of foundation models for other domains and data modalities such as images, code, and music. In this paper, we argue that business process data representations have unique characteristics that warrant the development of a new class of foundation models to handle tasks like process mining, optimization, and decision making. These models should also tackle the unique challenges of applying AI to business processes which include data scarcity, multi-modal representations, domain specific terminology, and privacy concerns.