building
AI Gateway Playground
A small lab for comparing model routing, fallback behavior, latency, and observability across LLM providers.
Projects / Evidence through building
Questions turned into working systems.
Small experiments are welcome here. Each project starts with a question and grows only when the evidence is useful.
exploring
Reliable LLM Experiments
Reproducible experiments for measuring LLM application quality, regressions, and failure modes.
exploring
Streaming Systems Lab
Notes and prototypes around Flink, Kafka, event time, provenance, and stream-processing reliability.