Honored to be part of this remarkable journey with the #RIOTULab at Prince Sultan University, where we achieved #3rd place among 170 global teams in the #ALLAM_Challenge! This achievement wouldn’t have been possible without the incredible leadership, vision, and unwavering support and trust of Prof. Anis Koubaa His guidance truly inspires us to push boundaries and excel in the field of AI. 🚀 Here’s to many more milestones ahead!
Generative AI Expert| Research and Innovation Expert & Consultant | CS Full Professor | Drone/UAV Expert & Conultant| Large Language Models | Robot Operating System | Arabic LLMs | RAG and Chatbots | AI and Data Science
🌟 𝗧𝗵𝗲 𝗪𝗵𝗼𝗹𝗲 𝗦𝘁𝗼𝗿𝘆: 𝗔 𝗣𝗿𝗼𝘂𝗱 𝗔𝗰𝗵𝗶𝗲𝘃𝗲𝗺𝗲𝗻𝘁 𝗳𝗼𝗿 𝗣𝗿𝗶𝗻𝗰𝗲 𝗦𝘂𝗹𝘁𝗮𝗻 𝗨𝗻𝗶𝘃𝗲𝗿𝘀𝗶𝘁𝘆’𝘀 𝗥𝗼𝗯𝗼𝘁𝗶𝗰𝘀 𝗮𝗻𝗱 𝗜𝗻𝘁𝗲𝗿𝗻𝗲𝘁-𝗼𝗳-𝗧𝗵𝗶𝗻𝗴𝘀 𝗟𝗮𝗯 𝗧𝗲𝗮𝗺! 🌟 𝗔𝗹𝗵𝗮𝗺𝗱𝘂𝗹𝗶𝗹𝗹𝗮𝗵, with immense gratitude, we are thrilled to share that the RIOTU Lab at Prince Sultan University has achieved 𝟯𝗿𝗱 𝗽𝗹𝗮𝗰𝗲 𝗮𝗺𝗼𝗻𝗴 𝟭𝟳𝟬 𝗰𝗮𝗿𝗲𝗳𝘂𝗹𝗹𝘆 𝘀𝗲𝗹𝗲𝗰𝘁𝗲𝗱 𝘁𝗲𝗮𝗺𝘀 𝗳𝗿𝗼𝗺 𝘁𝗵𝗼𝘂𝘀𝗮𝗻𝗱𝘀 𝗼𝗳 𝘁𝗲𝗮𝗺𝘀 around the world in the 𝗔𝗟𝗟𝗔𝗠 𝗖𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲! This competition, organized by the @𝗦𝗮𝘂𝗱i 𝗗𝗮𝘁𝗮 𝗮𝗻𝗱 𝗔𝗿𝘁𝗶𝗳𝗶𝗰𝗶𝗮𝗹 𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 𝗔𝘂𝘁𝗵𝗼𝗿𝗶𝘁𝘆 (SDAIA | سدايا) in partnership with the 𝗦𝗮𝘂𝗱𝗶 𝗙𝗲𝗱𝗲𝗿𝗮𝘁𝗶𝗼𝗻 𝗳𝗼𝗿 𝗖𝘆𝗯𝗲𝗿𝘀𝗲𝗰𝘂𝗿𝗶𝘁𝘆, 𝗣𝗿𝗼𝗴𝗿𝗮𝗺𝗺𝗶𝗻𝗴, 𝗮𝗻𝗱 𝗗𝗿𝗼𝗻𝗲𝘀, challenged teams to push the boundaries of AI-powered, real-time interactive experiences. Special thanks to Prince Sultan University for their support to the RIOTU Lab to make these achievements. 🤖𝘁𝗵𝗲 𝗺𝗼𝗺𝗲𝗻𝘁𝘀: • The instant of award winning announcement https://lnkd.in/d4dGVDaa • A Post Award Winning Interview https://lnkd.in/dXCp5x9H • Demo of the Interactive Chatbot https://lnkd.in/dkwRAfnZ 🚀 𝗣𝗿𝗼𝗷𝗲𝗰𝘁 𝗖𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲𝘀 Creating a truly interactive experience with ALLAM involved overcoming several technical hurdles, especially in real-time communication. Latencies from 3D rendering, text-to-speech, speech recognition, and LLM API calls were major sources of delay. Through our team’s dedication, we successfully reduced response times from tens of seconds to under 3 seconds, ensuring an uninterrupted, smooth user experience. 🔧 𝗞𝗲𝘆 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗲𝘀 • 𝗔𝗟𝗟𝗔𝗠 𝗔𝗣𝗜: Powered by a robust backend and hosted on Digital Ocean with 8 CPUs to maintain high speed and quality. • 𝗥𝗔𝗚 (𝗥𝗲𝘁𝗿𝗶𝗲𝘃𝗮𝗹-𝗔𝘂𝗴𝗺𝗲𝗻𝘁𝗲𝗱 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻): This advanced Arabic RAG model combines precise search with retrieval to provide accurate responses using a vector database, enhancing the relevance of results. • 𝗦𝗽𝗲𝗲𝗰𝗵 𝗮𝗻𝗱 𝗟𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴: The backend integrates sophisticated Arabic speech-to-text and text-to-speech conversion systems. • 𝗩𝗶𝗿𝘁𝘂𝗮𝗹 𝗖𝗵𝗮𝗿𝗮𝗰𝘁𝗲𝗿 𝗖𝘂𝘀𝘁𝗼𝗺𝗶𝘇𝗮𝘁𝗶𝗼𝗻: Users can customize their avatar to answer queries from ALLAM's data or personal data, allowing for an authentic and adaptive experience. 🌍 This milestone was achieved thanks to the collaborative effort of: • Dr. Adel Ammar • Dr. Bilel Benjdira • Engineer Omar Najjar • Engineer Anas Magdy • Engineer Hamad Al-Oqaily • Engineer Sultan Hamad
يا له من إنجاز رائع
Congrats OMER
GenAI Intern @ Spiro Carbon Group | BSc Student in Computer Science
2dCongratulations 🥰❤️❤️