This paper presents a novel approach to empower female founders by integrating a Large Language Model (LLM) into a serious game featuring player-generated content. The game leverages interactive visual novels to enhance resilience and awareness of gender-based discrimination in startup environments. The implementation of generative AI for bias detection in the conversations encountered by founders enables the provision of insightful feedback on discriminatory elements within the visual novels in which players immerse themselves. The game allows players to create personalized visual novels, reflecting their own entrepreneurial experiences. The system selects backgrounds, characters, and dialogue options, transforming players into creators while identifying and providing feedback on discriminatory elements within stories. The game's architecture, designed for modularity, supports the integration of various LLMs, enhancing its adaptability and future-proofing. Preliminary results demonstrate the game's feasibility, with expert evaluations affirming the effectiveness of its bias and discrimination detection mechanisms. This research contributes to the fields of gender equality and entrepreneurship in the digital domain, highlighting the potential of serious games and AI in addressing social challenges.