ericstj 

Pure managed HNSW approximate-nearest-neighbor search for .NET with SIMD-accelerated vector distances and no native dependency.

Managed, dependency-free port of Facebook fastText inference for text/language classification. Loads pretrained supervised models (e.g. lid.176.ftz) and predicts without any native dependency.

Pure-managed Model2Vec static embedding inference for .NET. Loads safetensors embeddings and Hugging Face tokenizers without native dependencies.

Managed .NET inference API for OpenAI Privacy Filter, using the official ONNX model with o200k_base tokenization and constrained BIOES decoding.

A fully-managed (no native dependency), AOT-friendly semantic + keyword search CLI for code and text. Combines Model2Vec static embeddings, Hnsw.Net vector search, BM25, and fastText language identification.

A fully-managed (no native dependency), AOT-friendly semantic + keyword search CLI for code and text. Combines Model2Vec static embeddings, Hnsw.Net vector search, BM25, and fastText language identification.

A fully-managed (no native dependency), AOT-friendly semantic + keyword search CLI for code and text. Combines Model2Vec static embeddings, Hnsw.Net vector search, BM25, and fastText language identification.

A fully-managed (no native dependency), AOT-friendly semantic + keyword search CLI for code and text. Combines Model2Vec static embeddings, Hnsw.Net vector search, BM25, and fastText language identification.

A fully-managed (no native dependency), AOT-friendly semantic + keyword search CLI for code and text. Combines Model2Vec static embeddings, Hnsw.Net vector search, BM25, and fastText language identification.