Sijan Paudel

I am a Generative AI and ML engineer specializing in production-grade LLM applications, MLOps, and backend development. I design and deploy RAG systems, QLoRA fine-tuning pipelines, multi-agent workflows, and GPU-accelerated inference for real-world use.

My work blends deep learning research with engineering rigor: I optimize model training and inference (FP16/BF16, INT8 quantization, TensorRT), and build reliable APIs and microservices with FastAPI and Django. I emphasize observability, caching, provider fallbacks, and cost-aware integrations.

Notable achievements include reducing RAG query latency by 30%, cutting fine-tuning time from 15h to 5h on constrained GPUs, and implementing distributed agent systems with monitoring and resilience. I also delivered backend projects with measurable performance and coverage improvements.

I mentor and teach cloud AI and deployment best practices, having led workshops and tutorials on Azure AI, GitHub, containerized deployments, and secure CI/CD. I seek roles where I can drive LLM productionization, MLOps excellence, and scalable AI product development.