Agents
Live demo
Predictive maintenance dashboard
Multi-agent monitoring for industrial equipment: anomaly detection, failure estimates, and a live dashboard you can open.
View projectAI infrastructure · Applied AI
Leonard Zhu
AI infrastructure & applied AI practitioner — building and explaining the stack from energy and chips to models and products.
Featured proof
Selected work
A few pieces that show the practice: a live multi-agent dashboard, edge AI on Jetson, and a small model pipeline with a public repo.
Edge
Jetson
Edge AI on NVIDIA Jetson
Running generative models locally on a Jetson Orin Nano, without sending every inference to the cloud.
View project
Multimodal
GitHub
Image-to-audio story
A small app that turns a picture into a spoken story with image-to-text, a language model, and text-to-speech.
View projectThe map
The AI industry stack
Five layers from energy to applications. Select a layer for the companies and the context behind it.
Learning taste
What I’m studying
Short notes on courses and books that connect infrastructure experience to applied AI. The full list lives on Learning.
NVIDIA AI infrastructure
Why it matters: deploying and operating GPU stacks is the bridge from enterprise infrastructure into AI systems.
See coursesHugging Face agents
Why it matters: tool use, planning, and multi-agent patterns show up in the maintenance dashboard work.
See coursesDesigning Data-Intensive Applications
Why it matters: reliability, scale, and maintainability are still the floor under every model layer.
See reading