Our project attempts to develop an AI-driven platform for neoantigen discovery and cancer immunotherapy. Our proprietary software ImmuNOVO will enable fast, sensitive and accurate me-antigen identification. We will then construct a data warehouse, from which sophisticated AI models will be trained for high throughput identification of neoantigens from cancer patients and for prediction of their immunogenicity. We expect the AI-driven approach will drastically accelerate the process, while reducing the cost of neoantigen discovery and custom cancer vaccine design.