Citation:

Mohammed Siddiq Hussain Nisar Khan, 2026. AI-Assisted Design-to-Code Automation for Enterprise Frontend Engineering,  Volume 6 Issue 3: 73-83.

Abstract:

The success of the latest developments in multimodal artificial intelligence and large language models has made Design-to-Code automation even more efficient and straightforward for converting a written design document into working front-end application source code. With the current explosion of multimodal AI and the advent of generative large language models, the above-mentioned Design-to-Code automation initiative takes on new momentum by turning text design docs into usable application front-end code. While there has been significant progress, such as visual fidelity and code generation accuracy, it is important to recognize that existing literature largely concentrates on maintaining visual fidelity and code generation accuracy with little attention given to software engineering requirements of real-world enterprise systems, such as Enterprise Design System compliance, continuous integration, security, or maintainability requirements. It is a comprehensive overview of the current research in creating novel automatic drawing techniques based on code, such as new developments in rule-based drawing, Deep Learning approaches to drawing, as well as Vision-Language models, Multimodal LLMs and Agentic AI. It's a methodology that has been developed based on experience with the approach PRISMA - Frontend Engineering in the Enterprise Application - which has been created to identify the available approaches, evaluation criteria, application scenarios and deployment challenges. Issues identified in the review for further work are the need for benchmarking, based on enterprise principles; governance mechanisms on board; support for reusable component architectures; and consideration of software quality over the long term. To address the above limitations, a conceptual framework called Enterprise Design-to-Code Automation Framework is proposed that addresses the drawbacks of AI-generated code automation, design systems, quality assurance, governance and DevOps principles.

References:

[1] Chang, Y., Wang, X., Wang, J., Wu, Y., Yang, L., Zhu, K., ... & Xie, X. (2024). A survey on evaluation of large language models. ACM transactions on intelligent systems and technology, 15(3), 1-45.

[2] Beltramelli, T. (2018, June). pix2code: Generating code from a graphical user interface screenshot. In Proceedings of the ACM SIGCHI symposium on engineering interactive computing systems (pp. 1-6).

[3] Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., ... & Polosukhin, I. (2017). Attention is all you need. Advances in neural information processing systems, 30.

[4] Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., ... & Houlsby, N. (2020). An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929.

[5] Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., ... & Sutskever, I. (2021, July). Learning transferable visual models from natural language supervision. In International Conference on Machine Learning (pp. 8748-8763). PmLR.

[6] Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., ... & Kiela, D. (2020). Retrieval-augmented generation for knowledge-intensive NLP tasks. Advances in neural information processing systems, 33, 9459-9474.

[7] Achiam, J., Adler, S., Agarwal, S., Ahmad, L., Akkaya, I., Aleman, F. L., ... & McGrew, B. (2023). Gpt-4 technical report. arXiv preprint arXiv:2303.08774.

[8] Si, C., Zhang, Y., Li, R., Yang, Z., Liu, R., & Yang, D. (2024). Design2code: Benchmarking multimodal code generation for automated front-end engineering. arXiv preprint arXiv:2403.03163.

[9] Cho, J., Kang, D., Kim, H., & Lee, G. G. (2025). Self-Correcting Code Generation Using Small Language Models. arXiv preprint arXiv:2505.23060.

[10] Li, Z., Wu, X., Du, H., Liu, F., Nghiem, H., & Shi, G. (2025). A survey of state-of-the-art large vision-language models: Benchmark evaluations and challenges. In Proceedings of the Computer Vision and Pattern Recognition Conference (pp. 1587-1606).

[11] Gülmez, B. (2026). Code generation with large language models: a survey from neural program synthesis to autonomous software development. Applied Intelligence, 56(6), 200.

[12] Tao, Y., Li, Y., Qin, Y., & Liu, Y. (2025). Retrieval-augmented code generation: A survey with focus on repository-level approaches. arXiv preprint arXiv:2510.04905.

[13] Acharya, M., Zhang, Y., Leach, K., & Huang, Y. (2025). Optimizing code runtime performance through context-aware retrieval-augmented generation. arXiv preprint arXiv:2501.16692.

[14] Lv, Z., Poiesi, F., Dong, Q., Lloret, J., & Song, H. (2022). Deep learning for intelligent human–computer interaction. Applied Sciences, 12(22), 11457.

[15] Ganesaraja, R., AP, S., Rathinasamy, K., Amancharla, C., Das, R., Panse, S. D., ... & Vijayakumar, S. (2025). Beyond Prototyping: Autonomous, Enterprise-Grade Frontend Development from Pixel to Production via a Specialized Multi-Agent Framework. arXiv preprint arXiv:2512.06046.

[16] Geruslu, V., Aliyeva, Z., & Tüzün, E. (2026). Factors Influencing the Quality of AI-Generated Code: A Synthesis of Empirical Evidence. arXiv preprint arXiv:2603.25146.

[17] Jain, V., Agrawal, P., Banga, S., Kapoor, R., & Gulyani, S. (2019). Sketch2Code: transformation of sketches to UI in real-time using deep neural network. arXiv preprint arXiv:1910.08930.

[18] Chen, J., Chen, C., Xing, Z., Xia, X., Zhu, L., Grundy, J., & Wang, J. (2020). Wireframe-based UI design search through image autoencoder. ACM Transactions on Software Engineering and Methodology (TOSEM), 29(3), 1-31.

[19] Jiang, Y., Zheng, Y., Wan, Y., Han, J., Wang, Q., Lyu, M. R., & Yue, X. (2025). Screencoder: Advancing visual-to-code generation for front-end automation via modular multimodal agents. arXiv preprint arXiv:2507.22827.

[20] Yue, C., Chai, J., Zhang, Y., Ding, Z., Liang, X., Wang, P., ... & Lin, W. (2025, November). UIOrchestra: Generating High-Fidelity Code from UI Designs with a Multi-agent System. In Findings of the Association for Computational Linguistics: EMNLP 2025 (pp. 2769-2782).

[21] Wang, X., Li, B., Song, Y., Xu, F. F., Tang, X., Zhuge, M., ... & Neubig, G. (2025, May). Openhands: An open platform for AI software developers as generalist agents. In International Conference on Learning Representations (Vol. 2025, pp. 65882-65919).

[22] Yang, Z., Hong, W., Xu, M., Fan, X., Wang, W., Cheng, J., ... & Tang, J. (2025). UI2Code^ N: A Visual Language Model for Test-Time Scalable Interactive UI-to-Code Generation. arXiv preprint arXiv:2511.08195.

Keywords:

AI-Assisted Design-to-Code, Enterprise Frontend Engineering, Large Language Models (LLMs), Multimodal AI, Software Engineering Automation