{ ai_agent }LangGraph FastAPI Groq Tavily Docker
Hi, I'm Kailash,
AI engineer & full‑stack developer
I build multi‑agent systems, RAG pipelines, and production LLM apps — from Figma to Docker. Every project ships with a real problem, a deliberate tech decision, and (mostly) a live URL.
Currently building Multi-Agent Systems
"
I own the full stack — Figma to Docker, schema to deployment. Every project ships with a real problem, a deliberate tech decision, and a live URL.
— Kailash
#projects
View all ~~>{ voice_ai_chat }FastAPI Groq Whisper LLaMA 3.3 Docker
{ rag_document_chat }LangChain ChromaDB Groq HuggingFace
{ scheduling_assistant }LangGraph Streamlit SQLite Groq
{ support_rag_agent }LangChain FAISS Streamlit Groq
Code / Experimentsno live deploy yet
{ charityconnect }React 18 TypeScript PostgreSQL Drizzle
{ exception_inbox }FastAPI React SQLAlchemy Docker
Exception Inbox
Plan‑vs‑actual production deficit tracker — ingest pipeline, exception detection, and a day‑by‑day review UI.
{ binance_futures_trading_bot }Python CLI REST API pytest
Binance Futures Trading Bot
CLI trading bot for Binance Futures Testnet — MARKET, LIMIT & STOP orders with validation, structured logging, and 25 unit/integration tests.
{ phaedrix }FastAPI Multi-Agent Python
Phaedrix
Supervisor-router multi-agent system coordinating Research, Code, Data & Utility agents behind a ChatGPT-style UI.
#skills
LangChain
LangGraph
HuggingFace// currently building
→ LangGraph multi-agent scheduling assistant (Triage + Booking agents) [shipped ✓] → Cited, memory-aware RAG support agent (LangChain + FAISS) [shipped ✓] → Streaming voice AI with conversation memory [shipped ✓] → Phaedrix — supervisor-router multi-agent system [in progress] → CI/CD pipelines across all projects [in progress]
#about-me
Hello, I'm Kailash —
I design and ship production AI systems end‑to‑end. From the schema and prompt graph to the React surface and the Docker image that runs in front of users — I own the whole stack.
I specialize in multi‑agent LLM architectures, retrieval pipelines with grounded citations, and real‑time voice systems with sub‑2s round trips. I obsess over latency budgets, evaluation harnesses, and the boring infra that makes models reliable in production.
Recent work: an autonomous agent with live web search and code execution, a streaming Whisper→LLaMA voice assistant, a LangGraph scheduling assistant that negotiates conflicts and persists state across sessions, and a cited RAG support agent with multi‑turn memory — most shipped with a live URL, all with clean engineering decisions behind them.
const kailash = {
role: "AI Engineer",
focus: "Multi-Agent + RAG",
stack: "end-to-end",
based: "Ghaziabad, IN",
available: true,
};
#contacts
I'm open to internships, freelance projects, and collaborations around AI engineering, full‑stack development, and product design. If you have a problem worth solving, let's talk.