Avinish Kumar

AI Engineer — Agents & Backend

Gurugram, Indiamr.avinish@gmail.comlinkedin.com/in/avinishkumargithub.com/techavinish

Summary

Backend engineer (4+ years) who builds AI agents. Software Engineer, AI Platform at Leena AI, Gurugram, India: the runtime behind autonomous AI agents for 500+ enterprises. Streaming LLM conversations over SSE/WebSockets, tool-calling agents on platform APIs, agent auth (JWT/ACL, RFC 7662 introspection), and A2A/MCP agent surfaces on Node.js/TypeScript microservices across multi-region AWS. Targeting Senior AI Engineer roles: AI agents, LLM systems and the backends behind them.

Skills

Backend
Node.js · TypeScript · Microservices · SSE / WebSockets · MongoDB · RabbitMQ · Kubernetes · Docker · AWS · Python
AI & LLM
Claude API · Claude Code · MCP · A2A · A2UI · LangChain & LangGraph · Tool-calling runtimes · Multi-agent pipelines · RAG & semantic search · Prompt engineering
Frontend
React · Next.js · Redux & React Query · Tailwind · Storybook
Practices
Distributed systems · Zero-downtime migrations · CI/CD & releases · Jest & testing · Auth: JWT, SAML, OAuth · L3 production support

Experience

Leena AI — Software Engineer, AI Platform

Jul 2022 - Present
  • Built the Agent Dashboard, an AI operator agent in the admin console that streams responses over SSE, executes tools against platform APIs, and lets admins configure and debug bots conversationally; exposed it over A2A and MCP so external agents can drive the same workflows.
  • Own the real-time conversation runtime for 500+ enterprises: SSE/WebSocket streaming, server-driven UI components, reconnect message replay, and agent token auth (JWT/ACL) on Node.js/TypeScript microservices across multi-region AWS.
  • Built Claude Code + MCP engineering automation adopted team-wide: a multi-agent PR review pipeline (60-80% lower LLM token spend per review), agentic E2E test generation, and custom MCP servers for semantic code search across 9 repos, cutting review and test cycles from hours to minutes.
  • Shipped 4 channel and auth integrations end to end: WhatsApp v2 revamp, MS Teams provisioning, live-chat handoff with RFC 7662 token introspection, and SAML SSO.
  • Lead L3 production support and release management for 3 product modules: built a mandatory pre-deploy sign-off gate (scalability, reliability, feature flags, latency), RabbitMQ service-level health checks, and zero-downtime DB migration patterns.
  • Led the type-safety migration across the Node.js/TypeScript monolith and microservices, enabling noImplicitAny and resolving 948 type errors.
  • Spearheaded a 30+ component design system and micro-frontend framework used by 25+ engineers; migrated legacy frontends to a modular monorepo, cutting build times 50% and improving web app load performance 35% via Lighthouse-driven optimization.

Cognizant — Software Engineer Intern

Feb 2022 - Jul 2022
  • Migrated a legacy BackboneJS frontend to ReactJS and shipped new Chrome-extension features.
  • Created end-to-end web test automation scripts, reducing manual QA effort and accelerating release cycles.

Education

Indian Institute of Technology, Patna

M.Tech, Artificial Intelligence & Data Science Engineering — 8.80 CPI

Jan 2024 - Dec 2025

Chandigarh Engineering College

Bachelor of Engineering, Computer Science — 7.80 CGPA

Aug 2018 - June 2022

MongoDB certified: Building AI Agents, AI-Powered Search with Vector Search, RAG Apps