Hello, I'm
Humayan Kabir
|
ML Engineer specializing in Deep Learning, LLMs & GenAI with 2+ years building scalable AI solutions and real-time data pipelines
Initializing Portfolio...
Hello, I'm
ML Engineer specializing in Deep Learning, LLMs & GenAI with 2+ years building scalable AI solutions and real-time data pipelines
I'm a Machine Learning Engineer with a passion for pushing the boundaries of AI research and building real-world applications that make an impact. I specialize in Deep Learning, LLMs, and GenAI Applications.
With a B.Sc. in Computer Science & Engineering from North South University and 3 years of professional experience in AI, Big Data, End to End deployment, and governance, I bring a strong foundation in both theory and practice. And I aim to utilize my skills to research and create innovative and helpful AI solutions.
My journey spans from developing Containerized Applications at an AgroTech, Big Data Solutions and High-availability Monitoring Systems at a Telco, Developing LLM powered AI solutions, to exploring cutting-edge research in symbolic regression and cryptographic analysis using neural networks.
Currently, I am working at a Law firm to develop useful and powerful LLM based AI solutions to optmize workflows. I thrive in collaborative, multicultural environments and have worked with teams across Asia, Europe, and North America.
Developed a RAG-based educational assistant for retrieval-driven feedback and smarter academic support. The aim was to provide personalized learning assistance for individuals. By highlighting student weakpoints, enhancing engagement and improving learning, not just grades.
Designed LLM-powered cloud pipelines to automate analytics workflows and streamline data operations. It's adaptable and scalable for various workloads.
AI Feynman is a published research work by researchers from MIT and Cornel University, among others. It uses Machine Learning to find new physical laws from data. This project extended the research by developing a transformer-based model using real and synthetic data.
An experimental project to understand the intersection of deep learning and SHA-256 cryptography. The goal was to do advance analysis of hash vulnerabilities and security. And it's implications for cybersecurity and data integrity.
Computer vision project choosing the most optimal model from CNN, RNN, and ResNet models. Trained and intended for both English and Bangla Sign Language recognition with live translation capabilities.
Enterprise-grade monitoring platform with parallel processing nodes achieving zero downtime. Additionally made 40% CPU optimization, and 50% storage reduction for telecom infrastructure.
I'm always interested in hearing about new opportunities, collaborations, or just connecting with fellow technologists.
Looking for roles that bridge AI research and real-world applications in global, culturally diverse teams.
Send an Email