Think it, Run it: Autonomous ML pipeline generation via self-healing multi-agent AI
The paper presents a multi-agent AI system that autonomously generates machine learning pipelines from datasets and natural-language goals. It introduces self-healing mechanisms, code-grounded retrieval, and adaptive learning to improve pipeline success rates and reduce development time. Evaluated on 150 ML tasks, the system achieves an 84.7% end-to-end success rate and outperforms baseline methods.
- ▪The system uses a five-agent architecture to handle data profiling, intent parsing, microservice recommendation, DAG construction, and execution.
- ▪It integrates code-grounded Retrieval-Augmented Generation (RAG) and an explainable hybrid recommender for improved microservice understanding and decision-making.
- ▪A self-healing mechanism powered by Large Language Models interprets errors and adapts based on execution history.
- ▪The approach was tested on 150 diverse ML tasks and achieved an 84.7% end-to-end pipeline success rate.
- ▪The study demonstrates that tightly coupled intelligent components in a unified architecture outperform isolated solutions.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2604.27096 (cs) [Submitted on 29 Apr 2026] Title:Think it, Run it: Autonomous ML pipeline generation via self-healing multi-agent AI Authors:Adela Bara, Gabriela Dobrita, Simona-Vasilica Oprea View a PDF of the paper titled Think it, Run it: Autonomous ML pipeline generation via self-healing multi-agent AI, by Adela Bara and 2 other authors View PDF Abstract:The purpose of our paper is to develop a unified multi-agent architecture that automates end-to-end machine learning (ML) pipeline generation from datasets and natural-language (NL) goals, improving efficiency, robustness and explainability.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv.org.
Discussion
0 commentsMore from arXiv.org
