Cloud Inference Orchestrator
A control plane prototype for routing AI workloads across regions and hardware classes.
A small systems prototype focused on cost-aware placement, observability, and predictable recovery.
Open-source tools and research prototypes
A control plane prototype for routing AI workloads across regions and hardware classes.
A small systems prototype focused on cost-aware placement, observability, and predictable recovery.
A compact observability dashboard for tracing AI requests across cloud services.
A tooling project for visualizing request latency, queue time, and retry behavior across distributed AI inference paths.
A searchable reading workflow for papers, notes, and recurring themes in cloud AI research.
A lightweight personal index that turns reading notes into reusable research memory.
project
project
An experiment in scheduling distributed AI jobs with cleaner resource isolation.
A systems prototype exploring how Ray workloads can be made easier to reason about in multi-tenant Kubernetes clusters.
A reproducible benchmark runner for multi-node model evaluation and regression tracking.
Built to make experiment drift visible before it becomes paper debt.
A lightweight personal dashboard for paper notes, tags, and reading momentum.
A minimal productivity surface for managing reading notes without turning the experience into a heavyweight knowledge base.