I am an AI Researcher with a strong track record of transforming cutting-edge research into robust, production-grade systems. I am actively seeking PhD opportunities and Research Scientist roles focused on Generative AI, Multimodal Representation Learning, and Multi-Agent Reasoning. I am deeply passionate about pushing the boundaries of knowledge-intensive reasoning while maintaining scalable and efficient model architectures.

• Multimodal Learning & Data Fusion: Integrating text, image, signal, and structured data for comprehensive context understanding.
• Agentic LLMs & Reasoning: Designing multi-agent orchestration frameworks for complex planning and multi-document summarization.
• Representation Learning: Exploring JEPA-style predictive architectures to generate compact, highly transferable semantic embeddings.
• Efficient AI (LLMOps): Optimization techniques including Parameter-Efficient Fine-Tuning (PEFT/LoRA), quantization (QLoRA/AWQ), and efficient inference serving.

• 🧪 Research Initiatives: Currently investigating multimodal data fusion methodologies and multi-agent coordination for knowledge-intensive reasoning. Innovating with predictive representation paradigms to enhance embedding robustness across diverse downstream tasks.
• ⚙️ Applied AI Engineering: Architecting and shipping highly scalable Agentic RAG pipelines.
  • Clinical Workflow Pilot: Developed an autonomous literature search & health advisory system achieving ~90% accuracy on proprietary datasets and a 4.0/5 user helpfulness rating.
  • Orchestration Systems: Enhanced cross-department throughput by ~200% through autonomous planning, LLM tool-use, and dynamic workflow automation.
• 🛠️ Infrastructure & Serving: Overseeing full-lifecycle deployability—from LoRA fine-tuning and quantization to highly optimized serving using vLLM, KV-cache management, and privacy-aware self-hosted architectures (Runpod/AWS).

I am always open to discussing:

• PhD Studentships / Research Assistantships in forward-thinking AI labs.
• Research collaborations aiming for top-tier conference publications (NeurIPS, ICLR, CVPR, ACL).
• Complex engineering challenges in GenAI, GraphRAG, and MLOps.