Greetings!

With over 10 years of experience in AI/ML research and development, I improve decentralised collaborative multi-agent, multi-robot, and multi-client AI/ML technologies to make them reliable under uncertainty, including uncertainty arising from distribution shift, heterogeneous collaboration, and asynchronous communication.

In collaboration with leading organisations such as CSIRO, I have a proven track record of improving multi-client AI systems by designing mathematically grounded deep learning pipelines to tackle drift. My recent work and collaborations focus on improving the reliability and scalability of LLM/VLM-based collaborative AI agents for multimodal data, including RAG systems. My work is supported by expertise in PyTorch, FastAPI, AWS, Docker, LangChain, LangGraph, LangSmith, MLOps practices, and HPC clusters for efficient multi-GPU training.

Across my career, I have built AI solutions across diverse data modalities, including text, tabular, sensor, image, CT scan, MRI, and LiDAR data, while tackling real-world challenges such as data imbalance, cold-start recommendation, explainability, and adversarial attacks. I enjoy teamwork in uncovering flaws in state-of-the-art algorithms, mathematical modelling, and developing proof-of-concept solutions to support algorithm design and decision-making. Beyond technical work, I love exploring the latest developments in AI/ML, data science, and software development. I contribute through mentoring, project supervision, and cross-sector research collaboration, with a commitment to reliable AI and real-world impact.