Bangladeshi technologist Mohaimen Khan has earned his second patent from Microsoft for a groundbreaking approach in Generative AI. His patented method is designed to build code-based context dynamically, improving accuracy and reducing hallucinations in Large Language Models (LLMs) — a key challenge in AI today.
Khan developed this innovative technique over the past year, aiming to refine how AI organizes and retrieves contextual information, resulting in more accurate and dependable responses. Sharing his journey on social media, he expressed deep gratitude for the recognition, stating, “I am incredibly grateful for my second patent on dynamically constructing code-based context for Generative AI.” 
Reflecting on his school days, Khan shared that he had often struggled with comprehending large volumes of information, a challenge that inspired his work in making information more accessible through AI. “I was never academically gifted and found it hard to process large amounts of data during elementary and middle school,” he noted.
Khan extended heartfelt thanks to his family, mentors, and supporters, adding, “Alhamdulillah, I am eternally grateful to Allah SWT for this opportunity.” His appreciation for Microsoft’s tools and platform underlined the supportive role technology can play in driving innovation.
This achievement shines a spotlight on Bangladesh’s growing influence in AI research, illustrating the country’s expanding contributions to global technological advancement.
