With the increase in demand of IT professionals, the employment opportunities for IT graduates are evolving. While demand for software developers and engineers is steadily rising, entrepreneurship is also an increasingly popular option for IT graduates. This paper proposes an optimization model for employment and entrepreneurship guidance for IT graduates using convolution neural network technology. A survey of IT graduates was conducted to understand needs and challenges around employment and entrepreneurship. Based on the survey results, an optimized credible neural network model was developed to provide personalized guidance on employment and entrepreneurship. The model utilizes deep learning techniques to analyze career preferences, skills, interests and other traits of graduates. It then matches graduates to suitable job openings, entrepreneurial opportunities, training programs and networking events. The network continuously learns from user interactions and feedback to improve recommendations over time. The recommendations are incorporated into the proposed model for optimization. Experimental results demonstrate that the optimized credible neural network model achieves better performance compared to baseline model without optimization. The proposed approach enables universities to better support IT graduates in navigating the evolving employment landscape by focussing on entrepreneurial activities and showing the impact of neural networks in understanding the employability index for IT graduates.