Emerging Multi-AI Agent Framework for Autonomous Agentic AI Solution Optimization

Kamer Ali Yuksel, Hassan Sawaf

Keywords: Agentic AI Systems, Multi-Agent AI Optimization, Iterative Refinement, AI-driven Hypothesis Generation, AI System Evaluation and Feedback, Automated AI System Adaptation, Self-Improving AI Agents, Adaptive AI Architectures, AI for Scientific Discovery, Autonomous AI Systems, AI-driven Experimentation, Self-Evaluating AI, AI Benchmarking & Standardization, AI-generated Hypothesis Validation, LLM-driven AI Optimization, Multi-Agent Coordination & Collaboration, Evolutionary AI Architectures, AI-driven Workflow Optimization, Trustworthy AI Systems, AI Model Transparency & Interpretability, AI Robustness & Error Mitigation, Human-in-the-loop AI for Science, AI Safety & Reliability

TL;DR: An emerging multi-agent framework that autonomously optimizes Agentic AI solutions using feedback loops, hypothesis generation, and automated refinement to enhance adaptability and performance with minimal human input.

Abstract: Agentic AI systems automate complex workflows but require extensive manual tuning. This paper presents a framework for autonomously optimizing Agentic AI solutions across industries, such as NLG-driven enterprise applications. It employs agents for Refinement, Execution, Evaluation, Modification, and Documentation, using iterative feedback loops powered by an LLM (Llama 3.2-3B). The system optimizes configurations without human input by autonomously generating and testing hypotheses, enhancing scalability and adaptability. Case studies demonstrate a significant boost in output quality, relevance, and actionability. Data, including original and evolved agent codes and outputs, are open-sourced.

Submission Number: 5