Pandas Studio

LangChain, LangGraph, RAG, Agent, vLLM, llama.cpp

Running LLM agents entirely on local models: vLLM and llama.cpp serving Qwen and Gemma on an NVIDIA DGX Spark, inside NemoClaw/OpenShell sandboxes with Hermes Agent.

AI training for companies and universities, LLM/RAG development and consulting, and conversations about LangGraph agents or running LLMs locally.