Headline

Experience in designing and optimizing production-grade RAG and agentic AI applications. Passionate about building efficient, accurate, and secure AI solutions that drive business impact.

About

I am an AI Engineer with over 12 years of experience in Software Engineering and 2+ year of specialized expertise in Generative AI and Agentic Systems. My core focus lies in designing and building production-grade Agentic AI applications that combine intelligence with action. I specialize in creating systems that don’t just answer questions — they understand, reason, and execute complex workflows autonomously. Key Expertise & Recent Work: Architected a complete Agentic AI Sales Agent using LangGraph for multi-step orchestration, incorporating Retrieval-Augmented Generation (RAG) and Model Context Protocol (MCP). Designed and implemented a hybrid RAG + MCP architecture with a dedicated MCP Server, enabling real-time actions such as live network status checks, security alert monitoring, and automated ticket creation. Built robust Guardrails (input, output, and topic-based) along with a comprehensive evaluation harness to ensure safety, reliability, and low hallucination rates (achieved 0.91 Faithfulness score). Optimized the system for performance, achieving 30% improvement in response speed through model optimization and efficient chunking strategies. Integrated Arize Phoenix for full observability and tracing of agent decisions, tool calls, and generation steps. Developed the entire solution with a forward-looking hybrid architecture (local Ollama + Azure OpenAI readiness) to ensure scalability and enterprise compliance. I thrive at the intersection of LLM engineering, agentic workflows, and production deployment. My approach combines deep technical rigor with a strong focus on business impact — whether it’s accelerating sales cycles, enabling intelligent operations, or building scalable AI platforms. Currently passionate about multi-agent systems, LLMOps, LangGraph orchestration, and enterprise AI adoption on cloud platforms like Azure. Open to discussions on Agentic AI, RAG + MCP architectures, or collaboration opportunities in Generative AI.

Location

Bengaluru

AI stack

codex

grok

hugging face

langchain

langgraph

langsmith

Work experience

AI Engineer at Confidential, Jun 2023 - Present · 3 y 3 mo, Bengaluru, Developed a production-ready Agentic AI Sales Agent that intelligently combines RAG (Retrieval-Augmented Generation), MCP (Model Context Protocol), and LangGraph orchestration. The agent autonomously qualifies leads, retrieves relevant knowledge from company data, performs real-time technical validation (network status, security alerts), and generates personalized proposals. Built with Guardrails, Phoenix observability, and hybrid architecture (local + Azure OpenAI ready), achieving high faithfulness (0.91) and 30% faster response times. This solution transforms traditional pre-sales processes into intelligent, automated workflows for faster deal closure and improved sales efficiency.

Android Developer at Confidential, Jun 2021 - Jun 2025 · 4 yrs 1 mo

Senior Software Engineer at Rakuten, Dec 2019 - Nov 2021 · 2 yrs, Bengaluru Area, India

Senior Software Engineer at HappyEMI, Apr 2017 - Feb 2019 · 1 yr 11 mos, Bangalore

Senior Software Engineer at CodeCraft Technologies Private Limited, Jul 2016 - Mar 2017 · 9 mos, Bengaluru Area, India, 1. Work with Office 365 REST APIs powered by Microsoft Graph into our Application to connect files.

Senior Software Engineer at BMS INNOLABS PVT LTD, Nov 2015 - Jul 2016 · 9 mos, Bengaluru Area, India, 1. Work with Payment Gateways and In-app Billing 2. Structure the App with MVP Pattern. 3. Enhance Dagger and RxJava Skills

Education

Bachelor of Engineering (B.E.), Computer Science, Reva Institute For Engineering And Technology Studies

12th, PCMB, Sri Vidya Mandir PU College

Public links

github: https://github.com/kgvinay