I am a Machine Learning Engineer at Beam AI building production-grade AI — LLM evaluation frameworks, self-healing pipelines, RAG systems, and autonomous agents. Concurrently, I serve as a Research Intern at the Empathic Computing Laboratory, University of Auckland, developing multimodal distributed LLM systems that integrate real-time neural and physiological signals for emotionally intelligent human–AI interaction.
My work is grounded in the conviction that reliable, interpretable, and multimodal AI is not a future ambition — it is an engineering problem solvable today.
- LLM Engineering: RAG systems, agentic workflows, evaluation frameworks, prompt engineering, and MLOps for production AI.
- Multimodal AI: Fusing language, vision, and physiological signals for intelligent human–AI interaction.
- Computer Vision: Real-time detection, segmentation, and understanding for robotics and automation.
- Robotics: SLAM, path planning (ROS2), and intelligent control systems.
- Open Source: Contributing to impactful AI/ML tools and frameworks.
- Multimodal LLM Architectures: Distributed systems for emotionally intelligent conversational agents (real-time EEG + physiological signal integration).
- LLM Reliability: Self-healing frameworks, LLM-as-a-Judge evaluation, deterministic output control.
- Advanced RAG: Multi-stage semantic chunking, hybrid retrieval, and domain-specific fine-tuning.
- MLOps at Scale: Model monitoring, anomaly detection, and automated evaluation pipelines on Azure & AWS Bedrock.
- LLM / AI Research: RAG, evaluation frameworks, agentic systems, multimodal AI.
- Robotics Systems: Autonomous navigation, control, and perception pipelines (ROS2).
- Open Source Initiatives: LLM tooling, model evaluation, agent frameworks.
- STEM Outreach & Mentorship: Supporting aspiring developers and researchers.
I was selected as the research intern from Pakistan at the Empathic Computing Lab — a leading international research group in AI, ML, and HCI. I completed my Bachelor's degree on a 100% merit-based scholarship for all 4 years at Habib University. I'm deeply inspired by Boston Dynamics' robotics vision and aim to advance robotic intelligence for positive human impact. When not building systems, I enjoy mentoring students and contributing to open-source tools that make AI more accessible and responsible.
