This project is a proof-of-concept AI-powered drug discovery platform focused on identifying potential drug candidates for neglected Indian diseases such as dengue and drug-resistant tuberculosis using only computational methods. The goal is to demonstrate that artificial intelligence and publicly available biomedical data can significantly speed up early-stage drug discovery at low cost, without involving any laboratory or clinical work. The work involves collecting and cleaning public chemical and bioactivity datasets, preparing molecular representations (such as SMILES, fingerprints, or descriptors), and building baseline machine-learning models to predict drug–target interactions. The freelancer will perform virtual screening and basic molecular docking on a limited set of compounds, rank potential drug candidates based on predicted effectiveness and basic safety or drug-likeness indicators, and document the entire pipeline clearly. All work must be reproducible, Python-based, and suitable for future extension when additional funding or lab partnerships are available. The final outcome is a working in-silico pipeline, cleaned datasets, a ranked list of promising drug candidates, and a short technical summary explaining the approach and results. This PoC is intended to validate technical feasibility and support future research funding, not to produce a final medicine or conduct experiments on patients or in labs.