The ScienceGPT project is a remarkable example of technological progress in the realm of scientific research, specifically in the field of material science. At the core of this initiative lies Llama2, an advanced open-weight Large Language Model (LLM) that has been carefully calibrated to cater to the scientific community's needs. One of the model's most noteworthy features is its self-defined vector store, constructed from a vast corpus of scientific literature and refined using advanced topic modeling techniques. This vector store significantly enhances the model's accuracy, resulting in a system that can generate and comprehend information with depth and precision firmly rooted in scientifically validated data. Additionally, the ScienceGPT project has introduced a revolutionary paradigm of scientific workflow, capable of generating simulation protocols that can be transformed into digital workflows using the SimStack framework. This approach bridges the gap between complex simulation protocols and scientists who lack a deep computational background, thus democratizing access to these protocols and making them more widely accessible.