What question did this study set out to answer?

The aim is to develop a multi-agent system for automating the synthesis of machine learning pipelines, with a focus on data privacy.

March 29, 2026Open Access

A Multi-Agent Architecture for Automated Machine Learning Pipeline Synthesis with Adaptive Control

Key Points

The aim is to develop a multi-agent system for automating the synthesis of machine learning pipelines, with a focus on data privacy.
Introduced MLAgents for end-to-end ML model synthesis.
Implemented a closed-loop control architecture with LLM agents for local execution.
Developed a non-LLM Improvement Agent to modify pipelines seen as a multi-armed bandit problem.
Utilized Thompson sampling for balancing exploration-exploitation in pipeline adjustments.
Achieved competitive performance through the adaptive policy.
Maintained strict privacy-by-design throughout the pipeline generation process.

Abstract

This paper introduces MLAgents, a multi-agent system for end-to-end ML model synthesis, emphasizing data privacy and adaptive optimization. In its closed-loop control architecture, LLM agents generate pipelines for secure local execution, protecting sensitive data. The core contribution is a non-LLM Improvement Agent that treats pipeline modification as a multi-armed bandit problem. This agent uses Thompson sampling, a Bayesian adaptive policy, to balance exploration-exploitation based on empirical feedback. We demonstrate that this principled approach achieves competitive performance while adhering to strict privacy-by-design principles.

A Multi-Agent Architecture for Automated Machine Learning Pipeline Synthesis with Adaptive Control

Key Points

Abstract

Cite This Study