Lazy evaluation, also known as call by need, is an evaluation strategy which defers all computations until a result is actually required. This is the opposite of eager evaluation, what Python normally does, where functions are executed as seen by the interpreter. Lazy evaluation gives the executor more context about how any given expression will be used, opening the door for interesting optimizations like intermediate object sharing or parallel execution. Python has some tools for emulating lazy evaluation, namely closures and generators; however, these objects cannot be used interchangably with non-lazy values. `lazy_python` is a library which allows users to automatically translate standard, eager Python functions into a lazy equivalent. This allows users to write or use standard Python code while getting the benifits of lazy evaluation. This poster will cover different ways to implement lazy evaluiation, focusing on the techniques used in `lazy_python`. It will show how to use the added execution context to implement optimizations. Finally, we will show a real example of using `lazy_python` with `dask` to automatically parallelize the execution of Python programs.