This study proposes a predictive model for evaluating corporate cultural characteristics and assessing cultural similarity between companies using publicly available corporate review data. In the context of mergers and acquisitions (M&A), cultural fit has been recognized as a critical factor influencing post-merger success. However, cultural evaluations are often constrained by limited access to internal company data. To address this challenge, we fine-tune open- and closed-sourced large language models (LLMs) with Parameter-Efficient Fine-Tuning (PEFT) techniques, specifically Low-Rank Adaptation (LoRA), to analyze employee reviews and generate assessments aligned with the Denison & Ko Organizational Culture Framework. The model predicts cultural characteristics across four key traits—Mission, Consistency, Involvement, and Adaptability—and their twelve sub-dimensions, providing both textual evaluations and quantitative ratings. Our results demonstrate that fine-tuned GPT-4o outperforms other models in generating accurate cultural evaluations and ratings. This research contributes to improving the objectivity and scalability of cultural fit assessments in M&A processes and highlights the potential of LLMs in extracting actionable insights from unstructured corporate review data.