Data modeling is the act of taking business requirements about a company or organization, it's processes, and the necessary information to make them run, and turning them into structured storage. This storage could be in the forms of normalized databases using Codd-style modeling rules usually in support of online transaction processing systems. It could be snowflake or star schema's in support of reporting or data warehouses. Data modeling may also apply to modeling the flow of data within the organization, between applications, etc. The focus is on accurately portraying the information that needs to be maintained in a way that everyone involved in the project, the business, the data professionals, and programmers can understand it. The requirements of your project dictate the types of models you're going to build and possibly how they get built. There's not really much more of a definition I can provide here. If you really don't know what a data model is or what might be applicable to your organization or project, it's probably time to hire a consultant to help you out rather than try to meet a requirement like this with minimal understanding. Data modelling is usually a very senior-level task.