Abdul Haque

Background

About Abdul Haque

Portfolio: http://abdul-haque-dev.surge.sh/I'm a machine learning practitioner with a strong foundation in deep learning and a growing fascination for causal reasoning and agentic AI. Over the years, I've worked across a wide range of domains from computer vision to generative models to language understanding always with the goal of building systems that not only learn but reason.My journey began with hands-on experience training and fine-tuning models like YOLOv8 for object detection and Stable Diffusion for generative tasks on custom datasets. I’ve built convolutional and recurrent neural networks from scratch for tasks like emotion recognition and video prediction, and developed time-series forecasting models using both LSTMs and feedforward networks. Along the way, I dove deep into GANs and DCGANs, pushing myself to understand the intricacies of training stable generative models.What truly excites me now is the intersection of causal inference and large language models. My final year project focuses on enhancing LLM reasoning through a graph-based RAG approach GraphRAG that provides structured causal context for more accurate understanding of news headlines, especially in the domain of fake news detection. I’ve built multi-agent workflows that integrate causal inference with intent detection, deploying these systems on Google Cloud Run for scalable, real-time access.I’m also deeply enthusiastic about agentic AI. I’ve spent considerable time exploring platforms like CrewAI, and have leveraged Agentic RAG techniques to build intelligent, autonomous agents capable of collaborating and reasoning over complex tasks. This passion for agentic systems stems from a desire to move beyond traditional task-specific AI toward more general-purpose, self-directed problem solvers.In addition to my AI work, I have strong web scraping experience using Selenium, BeautifulSoup, and REST-based APIs, and I’m confident in full-stack development and deployment. Whether it’s crafting a model, designing an architecture, or deploying a system to production, I enjoy every step of the process. I'm always looking to learn, collaborate, and push the boundaries of what intelligent systems can do.