This project is building a multi-agent system for autonomous scientific research discovery, leveraging large language models (LLMs) to streamline the research pipeline from data scraping to proposal generation. BeeARD is an agent-based system developed for autonomous research discovery, leveraging large language models (LLMs) to streamline the entire research pipeline—from data collection to proposal generation. Mirroring the foraging behavior of honeybees, BeeARD agents gather information from digital sources, compiling it into a central repository. This process transforms raw data into valuable knowledge, which can be utilized to generate novel research questions and optimize the assessment, funding, and impact of scientific research. Initially focused on psychedelic science for PsyDAO, this modular system is adaptable for all scientific research-focused bioDAOs. Key components include continuous data scraping, knowledge graph curation, lead generation tools, and proposal development. BeeARD agents will evolve through reinforcement learning, memory augmentation, human-in-the-loop (HITL), and fine-tuning, eventually operating autonomously with minimal human input. In other words, BeeARD agents are AI entities that live as digital brain cells of the DAO, turning spontaneous midnight thoughts and ephemeral musings into validated scientific possibilities. Within the DAO's digital ecosystem, these agents thrive, collaborate, and grow together, transforming fleeting insights into structured research, knowledge graphs, and actionable proposals. As part of the DAO's living network, they embody a collective intelligence, bridging the gap between raw inspiration and tangible scientific discovery.