Physics-based Traffic Control

Rate limiting for Node and the web: GCRA-default, one transform across in-memory/Redis/Postgres (proven bit-identical), and two-tier token leasing with a formally-verified overshoot bound independent of fleet size. Sync API, multi-dimensional checks, fixed-memory DDoS sketches.

Adaptive Rust pipeline for high-throughput file processing—dynamic chunking, parallelism, AES/ChaCha encryption, backpressure, and Prometheus/tracing.

Load limiting for JavaScript: adaptive concurrency limiting and slow-call circuit breaking. Sheds load when an upstream slows, not just when it errors.

An API gateway that decides which requests to fail under overload, and an experiment on whether adaptive concurrency control earns its keep

A zero-config, adaptive concurrency limiter for Spring Boot. Built on Java 21 Virtual Threads and TCP Vegas algorithm.

Adaptive process-local in-flight concurrency limiting from latency, throughput, and overload signals for Go.

High-performance adaptive distributed rate limiting API gateway powered by Redis Lua, two-tier in-memory leasing, and graceful degradation.

Adaptive concurrency limiting for .NET — a port of Netflix/concurrency-limits. TCP-congestion-style algorithms (Vegas, Gradient, AIMD) with ASP.NET Core middleware and gRPC interceptors.