Researchers at Seoul National University and Sungkyunkwan University in South Korea have developed an artificial intelligence platform that revolutionizes the fabrication of quantum-dot light-emitting diodes, or QLEDs. Led by Professor Jeonghun Kwak of Seoul National University and Professor Jaehoon Lim of Sungkyunkwan University, the team created a data-driven inverse design system that identifies optimal solvent conditions for depositing quantum dots, eliminating the need for conventional trial-and-error experimentation. The findings were published online on July 15 in Reports on Progress in Physics. QLED technology relies on solution-based coating processes, making it cost-effective for large-area displays. However, performance hinges on achieving uniform, dense quantum dot packing within the emitted film, a process heavily influenced by solvent properties. Historically, identifying the correct solvent composition required extensive manual testing due to the complex relationship between chemical properties and film morphology. To solve this, the research team trained a machine learning model on solvent characteristics such as vapor pressure, viscosity, density, and dielectric constant, paired with surface uniformity data obtained through atomic force microscopy. The AI model learned to inversely predict the exact solvent profile required to generate the most optimal quantum dot arrangement. While no commercially available single solvent matched the AI recommendations, the team successfully synthesized a custom solvent blend that precisely matched the predicted parameters. When implemented in actual QLED fabrication, this AI-optimized solvent blend doubled device efficiency and extended operational lifetime by more than forty times compared to conventional single-solvent methods. The breakthrough addresses a critical bottleneck in next-generation display manufacturing, significantly reducing development time and material waste. Supported by South Korea's Ministry of Science and ICT and the National Research Foundation through the Future Display Leading Technology and Nano & Material Technology Development Programs, this platform demonstrates the viability of AI-driven material science. Researchers anticipate the inverse design methodology will be rapidly adapted for other optoelectronic applications, including organic light-emitting diodes and photovoltaic cells, accelerating the commercialization of advanced electronic devices.