Build intelligent clients with MCP: From Query to Action
From function calling to the Model Context Protocol: how to build an MCP client that lets an LLM go beyond chat and take real actions through tools.
Vamsi Gutta · Engineering notes
Hands-on write-ups from 12 years of building software — from distributed systems to LLM-powered tools. Read a post; if you like it, there’s more about me below.
Latest post
From function calling to the Model Context Protocol: how to build an MCP client that lets an LLM go beyond chat and take real actions through tools.
Build a privacy-friendly Q&A bot over your own local documents using in-context learning, LangChain, FAISS vector search and a locally hosted LLM.
A step-by-step guide to building a chatbot in Python that answers FAQs from your product documentation.
About the author
Software Engineer · Platform Developer @ IBM Cloud
Software engineer with 12 years of experience designing and building backend services, shared libraries and distributed systems — currently developing platform software at IBM Cloud — plus full-stack and machine learning applications shipped to production.