Quality in, quality out. The growth in LLM and agent-based applications in drug discovery has heightened the importance of data quality, reproducibility, provenance, and validation. In this work, we present an agentic framework that leverages the data curation, QC, management, and multi-modal storage features in Paradigm4’s REVEAL platform. We demonstrate application of an LLM to query a knowledge base of summary statistics for validation of targets. Additionally, we discuss implementation of a broader framework to support the drug discovery pipeline, including agentic models for data harmonization, retrieval, patient journeys, among others.