In the framework of the international Cherenkov Telescope Array (CTA) gamma-ray observatory, the Italian National Institute for Astrophysics (INAF) is developing a dual-mirror, small-sized, end-to-end prototype (ASTRI SST-2M), inaugurated on September 2014 at Mt. Etna (Italy), and a mini-array composed of nine ASTRI telescopes, proposed to be installed at the southern CTA site. The ASTRI mini-array is a collaborative effort led by INAF and carried out by institutes from Italy, Brazil, and South-Africa. The project is also including the full data handling chain from raw data up to final scientific products. To this end, a dedicated software for the online/ on-site/off-site data reconstruction and scientific analysis is under development for both the ASTRI SST-2M prototype and mini-array. The software is designed following a modular approach in which each single component and the entire pipeline are developed in compliance with the CTA requirements. Data reduction is conceived to be run on parallel computing architectures, as multi-core CPUs and graphic accelerators (GPUs), and new hardware architectures based on low-power consumption processors (e.g. ARM). The software components are coded in C++/Python/CUDA and wrapped by efficient pipelines written in Python. The final scientific products are then achieved by means of either science tools currently being used in the CTA Consortium (e.g. ctools) or specifically developed ones. In this contribution, we present the framework and the main software components of the ASTRI SST-2M prototype and mini-array data reconstruction and scientific analysis software package, and report the status of its development.