With the advent of Large Language Models (LLMs), LLM-based agents are achieving greater autonomy through advanced agentic workflows and tool utilization. These agents have been deployed across various major domains, including software engineering, healthcare, and advanced research. Exploring the application and enhancement of these agents in diverse scenarios is becoming increasingly significant. This project is highly research-oriented and requires applicants to have a strong academic background, preferably with prior research experience in ML/NLP.