From Text Summarization to Task Automation: Exploring the Intersection of Large Language Models and Intelligent Agents

The emergence of Large Language Models (LLMs) has transformed natural language processing (NLP), enabling advanced capabilities in text comprehension, generation, and task automation. This paper explores the integration of LLMs into intelligent agent frameworks, emphasizing their role in autonomous decision-making across multiple domains. Through a literature review and detailed case studies, the research examines how LLMs—such as GPT, BERT, T5, LLaMA, and DeepSeek-R1—serve as foundational components for autonomous systems. A unified framework is proposed, comprising profiling, memory, planning, and action modules, to guide the architectural design of LLM-based agents. Case studies across medicine, statistics, software engineering, geography, finance, and art demonstrate the functional adaptability and domain-specific reasoning capabilities of these agents. Applications include clinical decision support, automated data science pipelines, autonomous coding systems, GIS data retrieval, real-time trading agents, and multimodal creative agents. The paper highlights the technical challenges of scale, hallucination, multimodal integration, and ethical concerns, while pointing to future directions such as personalized AI and explainable decision-making. By bridging the gap between LLM architectures and intelligent agent design, this study contributes to the development of more robust, scalable, and ethically aligned AI systems capable of real-world task execution.