Generative AI Engineer | LLMs, RAG & Multi-Agent Systems | Machine Learning & Deep Learning | AWS (EC2, ECS, Lambda) | FastAPI | CI/CD | Scalable AI & Backend Systems
Oklahoma City, United States
#OpenToWorkResearch & Coding and development interests include signal processing (SP), generative computer vision (CV) models, fine tuning of large language models (LLMs) for text comprehension, multivariate time series analysis, machine learning (ML), and deep learning (DL), in general. Research interests include signal processing, statistics, machine learning (ML), inverse problems, low dimensional models including group sparse, and low-rank models. Compressive sensing, applied probability theory, and role of convex optimization in estimating low-dimensional signal models.
I’m seeking a role as an AI/ML or Generative AI Engineer where I can build scalable, production-ready AI systems. I specialize in LLMs, RAG, prompt engineering, and multi-agent architectures, with experience in fine-tuning (PEFT, LoRA), FastAPI, vector databases, and AWS. I’m also experienced in designing end-to-end AI pipelines, optimizing inference performance, and deploying solution using CI/CD workflows. I’m interested in NLP, generative AI, and automation systems. I prefer to work remotely
Qult AI
Founded a startup and developed an AI-based controller for plunger lift systems to optimize gas well operations using live pressure and production data. The solution enhanced gas production efficiency, minimized methane emissions, and reduced equipment wear. Bootstrapped the venture and secured $100K in investor funding.
University of Engineering and Technology, Lahore
Georgia Institute of Technology
University of Engineering and Technology, Lahore
Brown University
Massachusetts Institute of Technology (MIT)
Massachusetts Institute of Technology (MIT)
INTECH Process Automation
Led the Signal Intelligence Division to develop AI-based predictive analytics products for industrial machinery monitoring and fault diagnostics in IIoT settings. Designed solutions for fault classification, anomaly detection, root cause analysis, and fault propagation pathways, processing hundreds of time-series datasets with advanced models to address high-dimensional distribution challenges. Delivered accurate predictions of remaining useful life for vibrating machinery, enhancing operational reliability and efficiency.
Netsol Technologies
AI in Fintech: Led cross-functional teams of frontend, backend, DevOps, and AI engineers to develop innovative solutions, including credit underwriting systems, customized car financing products, customer churn prediction models, and explainable AI for credit decisioning. Provided technical vision and strategic guidance to align marketing and business development efforts with the organization’s AI capabilities, ensuring impactful positioning of AI-driven solutions in the market.
Tellus You Care
AI in Healthcare: Developed predictive models to process radar point cloud data for monitoring elderly patients without attendants, enabling real-time tracking of activities and detection of health hazards like falls, heart attacks, and irregular sleep patterns. Addressed challenges of noisy point cloud data caused by electromagnetic wave propagation and multipath effects. Conducted extensive ML model development and advanced feature engineering to detect vitals, bed positions, room dimensions, and sleep patterns, ensuring accurate and reliable health monitoring.
OpenAI
Generative AI: Prepared question-answer datasets to train large-language models (LLMs) and assigned rewards based on the quality of datasets generated by the models using detailed evaluation rubrics. Utilized fine-tuning techniques with reinforcement learning to enhance LLM performance, actively contributing to the technical aspects of training workflows, reward modeling, and iterative fine-tuning processes.
Savantz AI
Generative AI: Designed and deployed custom diffusion models for consistent character generation in videos and images, leveraging variants of Stable Diffusion, MidJourney, and DALL-E for visual storytelling from text narratives. Built a custom pipeline integrating tools like Instant ID, ControlNet, and other state-of-the-art add-ons to ensure character consistency across scenes. Enhanced functionality with features such as face aging to match story progression while maintaining character coherence. Reproduced and extended state-of-the-art methodologies to create a stable and consistent character generation system. Deployed models at scale on AWS EC2 using CI/CD pipelines, FastAPI, and request throughput management with custom Redis Celery-like frameworks for queue management. Developed APIs incorporating off-the-shelf image-to-video models to generate animated videos from text stories, enabling seamless text-to-visual storytelling.
GAS Inc.
Generative AI: Designed and deployed fine-tuned LLM models to extract structured information from unstructured audit reports. Leveraging AWS services such as EC2, ECS, Lambda, and Airflow, I ensured scalable deployment and automation. I optimized LLM inference speed using the Unsloth package and implemented parameter-efficient fine-tuning (PEFT) to enhance model performance. My work included developing multi-agent systems for structured information extraction, database traversal, and question answering, employing RAG techniques and vector databases. I also established robust CI/CD pipelines using GitHub workflows and managed team collaboration through JIRA. Leading a cross-functional team of backend, frontend, machine learning engineers, and DevOps professionals, I successfully delivered a comprehensive AI-driven audit system.
PhD candidate in Electrical and Computer Engineering
MS in Electrical Engineering
B.Sc. in Electrical Engineering
Government of Pakistan
Fulbright Program
Fulbright Program
Lahore Board
