Abubakar S.

Are you looking for a specialist in AI and machine learning to turn your creative concepts into reliable, scalable, and state-of-the-art solutions? I have ten years of experience and have worked on AI-driven projects for more than seven of those years. I am an expert in creating end-to-end systems with the newest developments in machine learning and generative artificial intelligence. 𝗪𝗵𝗮𝘁 𝗦𝗲𝘁𝘀 𝗠𝗲 𝗔𝗽𝗮𝗿𝘁 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜 𝗮𝗻𝗱 𝗟𝗟𝗠 𝗘𝘅𝗽𝗲𝗿𝘁𝗶𝘀𝗲 𝗟𝗟𝗠𝘀 𝗠𝗮𝘀𝘁𝗲𝗿𝘆: Deep experience working with industry-leading language models, including Jais 13B, Zephyr, Mistral, Falcon, Claude, Bloomz, and LLaMA. 𝗠𝘂𝗹𝘁𝗶𝗹𝗶𝗻𝗴𝘂𝗮𝗹 𝗦𝗼𝗹𝘂𝘁𝗶𝗼𝗻𝘀: Delivered sophisticated RAG-based systems supporting Arabic, Turkish, and other languages, ensuring global applicability and enhanced user engagement. 𝗙𝗶𝗻𝗲-𝗧𝘂𝗻𝗶𝗻𝗴 𝗮𝗻𝗱 𝗢𝗽𝘁𝗶𝗺𝗶𝘇𝗮𝘁𝗶𝗼𝗻: Fine-tuned multiple 7B and 13B LLMs using PEFT and QLoRA for specific domains, improving their performance and efficiency. 𝗘𝗻𝗱-𝘁𝗼-𝗘𝗻𝗱 𝗔𝗜 𝗮𝗻𝗱 𝗠𝗮𝗰𝗵𝗶𝗻𝗲 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝗘𝘅𝗽𝗲𝗿𝘁𝗶𝘀𝗲 𝗔𝗜-𝗣𝗼𝘄𝗲𝗿𝗲𝗱 𝗡𝗟𝗣: Designed summarization tools, sentiment analysis systems, and call analytics platforms tailored for business insights. 𝗖𝗼𝗺𝗽𝘂𝘁𝗲𝗿 𝗩𝗶𝘀𝗶𝗼𝗻: Built AI models for document analysis, medical image segmentation, and historical data extraction. 𝗠𝗟𝗢𝗽𝘀 𝗠𝗮𝘀𝘁𝗲𝗿𝘆: Developed robust pipelines for model training, fine-tuning, and deployment on AWS SageMaker and GCP Vertex AI. 𝗛𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁𝗲𝗱 𝗣𝗿𝗼𝗷𝗲𝗰𝘁𝘀 𝟏. 𝐌𝐮𝐥𝐭𝐢𝐥𝐢𝐧𝐠𝐮𝐚𝐥 𝐑𝐀𝐆 𝐂𝐡𝐚𝐭𝐛𝐨𝐭 𝐰𝐢𝐭𝐡 𝐒𝐮𝐦𝐦𝐚𝐫𝐢𝐳𝐚𝐭𝐢𝐨𝐧 (i) Built a multilingual chatbot integrating Jais, Claude, and Mistral, enhanced with RAG for precise and contextual replies. (ii) Added summarization capabilities for condensing long multilingual content, ensuring seamless comprehension for end users. 𝟐. 𝐀𝐈-𝐏𝐨𝐰𝐞𝐫𝐞𝐝 𝐒𝐮𝐦𝐦𝐚𝐫𝐢𝐳𝐚𝐭𝐢𝐨𝐧 𝐟𝐨𝐫 𝐅𝐢𝐧𝐚𝐧𝐜𝐢𝐚𝐥 𝐑𝐞𝐩𝐨𝐫𝐭𝐬 Designed a table recognition and analysis system paired with Zephyr-based summarization tools to condense lengthy reports into actionable insights for executives. 𝟑. 𝐅𝐢𝐧𝐞-𝐓𝐮𝐧𝐞𝐝 𝐍𝐋𝐏 𝐌𝐨𝐝𝐞𝐥𝐬 𝐟𝐨𝐫 𝐂𝐚𝐥𝐥 𝐀𝐧𝐚𝐥𝐲𝐭𝐢𝐜𝐬 Implemented LLaMA-based text classification and extraction models, enabling automated analysis of call center data for improved customer service metrics. 𝟒. 𝐇𝐢𝐬𝐭𝐨𝐫𝐢𝐜𝐚𝐥 𝐌𝐞𝐭𝐚𝐝𝐚𝐭𝐚 𝐑𝐞𝐜𝐨𝐠𝐧𝐢𝐭𝐢𝐨𝐧 Leveraged Mask R-CNN for extracting hand-drawn metadata from historical maps, automating archival and retrieval processes. 𝟓. 𝐋𝐮𝐧𝐠 𝐂𝐚𝐧𝐜𝐞𝐫 𝐃𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧 𝐟𝐫𝐨𝐦 𝐂𝐓 𝐒𝐜𝐚𝐧𝐬 Developed segmentation models using OpenCV and deep learning techniques to improve diagnostic accuracy in healthcare applications. 𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐈 𝐄𝐱𝐜𝐞𝐥 𝐈𝐧 𝐋𝐋𝐌𝐬: Jais, Zephyr, Mistral, Falcon, LLaMA, Claude, Bloomz, and Stable Diffusion. 𝐂𝐥𝐨𝐮𝐝 𝐏𝐥𝐚𝐭𝐟𝐨𝐫𝐦𝐬: AWS (SageMaker, Lambda, ECS), GCP (Vertex AI, Storage). 𝐏𝐫𝐨𝐠𝐫𝐚𝐦𝐦𝐢𝐧𝐠 𝐚𝐧𝐝 𝐋𝐢𝐛𝐫𝐚𝐫𝐢𝐞𝐬: Python, PyTorch, TensorFlow, LangChain, LlamaIndex, OpenCV. 𝐃𝐚𝐭𝐚 𝐌𝐚𝐧𝐚𝐠𝐞𝐦𝐞𝐧𝐭: Expertise in vector databases like Pinecone , Chroma DB and pgvector for optimized data handling. 𝐖𝐡𝐲 𝐖𝐨𝐫𝐤 𝐖𝐢𝐭𝐡 𝐌𝐞? I thrive on solving complex challenges with innovative AI-driven solutions. Whether you're launching a new product, optimizing existing systems, or exploring the power of Generative AI, I deliver exceptional quality, timely results, and measurable outcomes. Let’s bring your AI and Machine Learning projects to life. Reach out today to discuss how we can collaborate on your next big idea!