Groq + LangChain fast inference pipeline

Real-time AI agents, low-latency RAG, high-throughput chat apps with production-scale inference speeds (500+ tokens/sec).