Summary
Joshua Cahyono is an ML Research Engineer with four years of hands-on experience applying state-of-the-art generative models, robotics control, and interpretability techniques across industry and academic labs. Based in Singapore and graduating from NTU with a 4.7/5.0 in Artificial Intelligence, he has shipped production-facing tooling and platforms—from a 2,000+ user RAG-enabled exam forum to standardized LLM evaluation pipelines and contributions to vision-language model development. His work spans PyTorch-based custom U-Net/ViT/Swin architectures, diffusion planners for robotic motion in MuJoCo, and real-world RL applications including offline RL for healthcare, highlighting a rare mix of ML research and systems engineering. He’s led cross-functional teams in student and institute settings, built the first interpretability framework for IC hardware security segmentation, and competed successfully in ICRA BARN challenges using CBF and MPC strategies. Ambitious and curious, he’s increasingly focused on multi-agent systems and robotics with the goal of translating ML advances into tangible societal impact.
4 years of coding experience
2 years of employment as a software developer
Bachelor of Science - BS, Artificial Intelligence, 4.7/5.0, Bachelor of Science - BS, Artificial Intelligence, 4.7/5.0 at Nanyang Technological University Singapore