Happy Yadav — AI Researcher

Hi I'm an AI Engineer focused on real-time Voice AI, Generative AI systems, and scalable backend architectures. I'm Currently building low-latency voice-to-voice pipelines, telephony-integrated conversational agents, and multi-agent LLM systems used in production environments. I do have hands-on experience with STT–LLM–TTS pipelines, WebRTC and VoIP systems, multi-agent orchestration using frameworks like LangGraph, LangChain, Pipecat, Livekit , and cost-aware inference optimization across OpenAI, Gemini, Groq models. I’ve also designed and optimized RAG-based chatbots with semantic retrieval pipelines, embedding optimization, context compression strategies, and structured response control for high-accuracy domain-specific use cases. Alongside this, I’ve worked on MCP tool ecosystems, enabling dynamic tool routing and controlled agent-tool interaction layers. Built real-time WhatsApp conversational bots and automation systems, including workflow-driven travel management solutions operating fully within messaging environments, handling structured data capture, async orchestration, and production-grade reliability constraints. I work at the intersection of AI systems, distributed backend engineering, and real-world deployment constraints focusing on latency, cost, observability, and scale. I'm actively sharing learnings around Voice AI, LLM infrastructure, RAG systems, and production-grade AI system design.

Stackforce AI infers this person is a Backend-heavy Fullstack AI Engineer specializing in Voice AI and real-time systems.

Location: Gurugram, Haryana, India

Experience: 2 yrs 2 mos