Generative AI is increasingly being integrated into education, offering transformative possibilities for both teaching and learning and tailoring educational content to the needs of individual students. We have chosen to elaborate on questions lying in the intersection of economics, linguistics and innovative engineering, aiming to benefit from automation achievable using AI. In our previous research we established methodological approaches to testing ability of Large Language Models, in particular Google Gemini and OpenAI ChatGPT, to process figurative speech and tasks in mathematics and physics defined in plain English; the present work is a solution for modelling financial projections of technological startups and competitor search. The said projections are based on an array of variables such as research and development cost, operational expenses, monthly increase of the customer count, cost of customer acquisition and so on; the input data is provided partially in numeric values and partially in the form of linguistic variables, the latter are translated into numbers (or weights) which are configurable, and modelling is done. The output of modelling is processed by a large language model to obtain a textual description in English, or any other supported language. This solution shall be used in the training of students of computer science and automation programs to help them master innovative project management and improve their business development skills. At this stage our solution relies on API-based access to 2 commercially available pretrained language model, namely, Gemini version 2.0 and OpenAI version GPT-4o, at a later stage we are considering development of our own language model. Besides, since our students are not English native speakers, generative capabilities of large language models assist them with the language barrier.