Popular repositories Loading

1. echoflow-voice-pipeline echoflow-voice-pipeline Public

   Real-time streaming voice assistant pipeline: ASR → LLM → TTS. Sub-second latency, circuit breakers, multi-backend, fully observable.

   Python

2. production-grade-rag production-grade-rag Public

   Production-grade RAG pipeline: hybrid search (BM25+vector+RRF), cross-encoder reranking, SSE streaming, OpenTelemetry metrics, Langfuse tracing, LLM-as-Judge evaluation, and FastAPI. Dual OpenAI/An…

   Python

3. local-ai-assistant local-ai-assistant Public

   Private-by-design offline AI resume optimizer & mock interviewer — powered by local Ollama SLMs with Gemini fallback

   Python

4. rag-observability-layer rag-observability-layer Public

   Production-grade monitoring, tracing, and CI regression-gating for RAG pipelines

   Python

5. structured-extraction-ft structured-extraction-ft Public

   Production QLoRA SFT+DPO pipeline & FastAPI inference server for schema-validated JSON extraction — runs on 4 GB VRAM

   Python