NaCode Studios

Semantic cache for LLM calls that prevents false hits via multi-guard lexical checks, calibratable thresholds, shadow-mode evaluation, pluggable embedders/stores, typed streaming responses and observability.

#wasm, #spring-boot, #llm, #coroutines, #ai

Async, cancellation-aware client for the Qdrant vector database offering a type-safe DSL, pluggable REST or gRPC transport, lightweight footprint, and modules for migration, observability, and adapters.

#llm, #ktor, #kotlin-native, #coroutines