Yasiru Rangana
AI Architect and AI Engineer, Sydney
Yasiru Rangana is an AI Architect and AI Engineer based in Sydney, Australia. He designs AI systems end to end and builds them to run in production: LLM applications, RAG and knowledge graphs, multi-agent workflows, and the harnesses around them. He has more than ten years of experience as a software engineer, and he still treats AI work as software engineering first.
He is a doctoral researcher at Western Sydney University, working on small object detection in fast-moving video, and builds open-source tools for AI-assisted development.
Focus
- Generative AILLM applications for production, with evaluation built in.
- RAGRetrieval and knowledge graphs that answer from real sources, with citations.
- Agentic AIAgents, tool use, multi-agent workflows and the harnesses that run them.
- AI architectureEnd-to-end system design, from cloud to GPU servers, with security and data control built in.
Work
- Open sourceTools for AI-assisted development, and code merged into the RAG frameworks LightRAG and RAG-Anything.
- ResearchSmall object detection in fast-moving video. Papers at ICMV 2010 and IEEE ABC 2025.
- AboutBackground, stack, education, certifications and contact.
Writing
- Running Next-Concept Modeling in the browser
A Sudoku solver and a story writer that draft the whole answer, then refine it, running on your own device.