Principles for Building a Product Based on Generative Neural Network Models
( Pp. 142-153)
More about authors
Mikhail V. Romanov
Moscow Institute of Physics and Technology (National Research University)
Dolgoprudny, Russian Federation Vladimir A. Lopatin
Moscow Institute of Physics and Technology (National Research University), Dolgoprudny, Russian Federation Valeriy E. Krivtsov Cand. Sci. (Phys.-Math.), Associate Professor, Senior Researcher
Moscow Institute of Physics and Technology (National Research University)
Moscow, Russian Federation
Moscow Institute of Physics and Technology (National Research University)
Dolgoprudny, Russian Federation Vladimir A. Lopatin
Moscow Institute of Physics and Technology (National Research University), Dolgoprudny, Russian Federation Valeriy E. Krivtsov Cand. Sci. (Phys.-Math.), Associate Professor, Senior Researcher
Moscow Institute of Physics and Technology (National Research University)
Moscow, Russian Federation
Abstract:
The availability of generative models through cloud APIs has lowered the barrier to building products around them, but it has not removed the underlying engineering problem: the mere ability to call a generative model does not guarantee that a workable, scalable, and degradation-resistant product can be built around it. The article systematizes the principles of building such a product – from selecting and combining generative models and designing a multiplatform delivery architecture to the methodology of testing technical hypotheses and the metrics used to evaluate generation quality. Using the case of a multiplatform service that generates personalized visual content, the article shows how the sequential testing of technical hypotheses – moving from simple face-swap to controllable generation with a fine-tuned diffusion model, automatic enrichment of user prompts, an image-to-image generation feature, compute optimization, and a cross-platform architecture – enables the transition from a prototype to a product that remains stable under a multi-fold increase in load. The article separately examines the principles of quality assurance for generative output, technical infrastructure scaling, and the mitigation of risks specific to generative AI products: dependency on a single model provider, quality degradation under load, and the vulnerability of fixed pricing to fluctuations in inference cost.
How to Cite:
Romanov M.V., Lopatin V.A. and Krivtsov V.E. Principles of product development based on generative neural network models. Computational Nanotechnology. 13, 2 (2026), 142–153. DOI: 10.33693/2313-223X-2026-13-2-142-153. EDN: XZRAUU
Reference list:
Romanov M.V., Lopatin V.A. Cheap production, expensive attention: The economics of innovation and diffusion of generative AI in the personalized visual content market. Creative Economy. 2026. Vol. 20, No. 7. Pp. 1523–1540. (In Rus.)
Bai J., Bai S., Chu Y. et al. Qwen technical report // arXiv. 2023. DOI: 10.48550/arXiv.2309.16609
Hessel J., Holtzman A., Forbes M. et al. CLIPScore: A reference-free metric for image captioning evaluation // Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). 2021. Pp. 7514–7528. DOI: 10.18653/v1/2021.emnlp-main.595.
Ho J., Jain A., Abbeel P. Denoising diffusion probabilistic models // Advances in Neural Information Processing Systems 33 (NeurIPS 2020). DOI: 10.48550/arXiv.2006.11239.
Hu E.J., Shen Y., Wallis P. et al. LoRA: Low-rank adaptation of large language models // Proceedings of the International Conference on Learning Representations (ICLR). 2022. DOI: 10.48550/arXiv.2106.09685.
Radford A., Kim J.W., Hallacy C. et al. Learning transferable visual models from natural language supervision // Proceedings of the International Conference on Machine Learning (ICML). 2021. Vol. 139. Pp. 8748–8763. DOI: 10.48550/arXiv.2103.00020.
Sculley D., Holt G., Golovin D. et al. Hidden technical debt in machine learning systems // Advances in Neural Information Processing Systems (NeurIPS). 2015. Vol. 28. Pp. 2503–2511. DOI: 10.48550/arXiv.1512.04155.
Wang L., Xing X., Cheng Y. et al. PromptEnhancer: A simple approach to enhance text-to-image models via chain-of-thought prompt rewriting // arXiv preprint. 2025. DOI: 10.48550/arXiv.2509.04545.
Bai J., Bai S., Chu Y. et al. Qwen technical report // arXiv. 2023. DOI: 10.48550/arXiv.2309.16609
Hessel J., Holtzman A., Forbes M. et al. CLIPScore: A reference-free metric for image captioning evaluation // Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). 2021. Pp. 7514–7528. DOI: 10.18653/v1/2021.emnlp-main.595.
Ho J., Jain A., Abbeel P. Denoising diffusion probabilistic models // Advances in Neural Information Processing Systems 33 (NeurIPS 2020). DOI: 10.48550/arXiv.2006.11239.
Hu E.J., Shen Y., Wallis P. et al. LoRA: Low-rank adaptation of large language models // Proceedings of the International Conference on Learning Representations (ICLR). 2022. DOI: 10.48550/arXiv.2106.09685.
Radford A., Kim J.W., Hallacy C. et al. Learning transferable visual models from natural language supervision // Proceedings of the International Conference on Machine Learning (ICML). 2021. Vol. 139. Pp. 8748–8763. DOI: 10.48550/arXiv.2103.00020.
Sculley D., Holt G., Golovin D. et al. Hidden technical debt in machine learning systems // Advances in Neural Information Processing Systems (NeurIPS). 2015. Vol. 28. Pp. 2503–2511. DOI: 10.48550/arXiv.1512.04155.
Wang L., Xing X., Cheng Y. et al. PromptEnhancer: A simple approach to enhance text-to-image models via chain-of-thought prompt rewriting // arXiv preprint. 2025. DOI: 10.48550/arXiv.2509.04545.
Keywords:
generative artificial intelligence, diffusion models, model fine-tuning, LoRA, prompt engineering, software architecture, MLOps, generation quality evaluation, scaling.
