Agent Dashboard
An AI operator agent inside the admin console. Streams responses over SSE, executes tools against platform APIs, and lets teams configure and debug bots conversationally.
ClaudeSSENode.js
Currently: Core Platform @ Leena AI
Backends that run AI agents.
Backend engineer building AI agents at scale: distributed systems, real-time streaming, message queues, zero-downtime migrations, and multi-region AWS.
3+
years shipping backend
60-80%
cheaper code review with AI agents
10x
faster delivery with AI tooling
About
I work on the Core Platform at Leena AI, where the systems I build carry every conversation between an enterprise employee and an AI agent: the streaming runtime, the token auth, the channel integrations. When a message survives a pod restart or a dropped socket, that is my code doing its job quietly.
I build most of it with Claude at the center of my workflow. The review pipelines, test generators, and code search tools I wrote for myself now run for my whole team. Before this I was an SDET at Cognizant, which is why I still write code assuming someone will try to break it. Along the way I earned an M.Tech in AI and Data Science from IIT Patna.
How an agent turn works
What happens between a user pressing send and an AI agent's reply streaming back. Each bar is one stage of the pipeline I build and operate; they fire in the order shown.
trace: one agent turn
ws: message receivedthe user hits send
auth: session token (JWT refresh)verify who is asking
runtime: build context + promptassemble the conversation
llm: stream response tokensthe model thinks, token by token
tool: execute + return resultthe agent takes an action
sse: deliver to clientthe reply streams back, replayed if the socket drops
Selected work
2026
Multi-agent PR review pipeline
A tiered pipeline of review agents with adversarial verification. Cut review cost 60-80% and turnaround from hours to minutes.
Claude APITypeScript
Streaming conversation runtime
Real-time SSE and WebSocket backbone serving hundreds of enterprises: reconnect message replay, server-driven UI, and delivery reliability across multi-region AWS.
WebSocketsRabbitMQMongoDB
Enterprise channel integrations
Shipped WhatsApp v2, MS Teams provisioning, external live-chat handoff with RFC 7662 token introspection, and SAML SSO, end to end.
OAuthSAMLREST
Semantic code search MCP
Natural-language code search across 9 repositories, built on tree-sitter parsing, BM25, and vector embeddings.
MCPEmbeddings
Design system & micro-frontends
A 30+ component design system and micro-frontend framework used by 25+ engineers across hundreds of enterprise deployments; monorepo migration cut build times 50%.
ReactStorybook
Skills
Backend
- Node.js
- TypeScript
- MongoDB
- RabbitMQ
- Kubernetes
- Docker
- SSE / WebSockets
- Microservices
AI & LLM
- Claude API
- Claude Code
- MCP
- Multi-agent pipelines
- RAG & embeddings
- Agentic workflows
Frontend
- React
- Next.js
- Redux
- React Query
- Tailwind
- Motion
Experience
Jul 2022 - Present
Software Engineer, Core Platform at Leena AI
Own the conversation runtime and agent auth behind an enterprise AI agent platform. Led a 948-error type-safety migration, shipped WhatsApp, MS Teams, and live-chat integrations, and run L3 support and releases for 3 product modules.
Node.jsTypeScriptMongoDBRabbitMQKubernetesClaude
Feb 2022 - Jul 2022
Software Engineer Intern at Cognizant
Migrated a legacy BackboneJS frontend to React and built end-to-end test automation that cut manual QA cycles.
ReactJavaScriptSelenium
Education
2024 - 2025
Indian Institute of Technology, Patna
M.Tech, Artificial Intelligence & Data Science Engineering
2018 - 2022
Chandigarh Engineering College
Bachelor of Technology, Computer Science
Certifications
MongoDB certified: Building AI Agents, AI-Powered Search with Vector Search, RAG Apps
Writing
Ask my agent
An agent trained on my profile. Ask it anything about me: it answers questions, hands over my resume, and can take you to any section of this site.
agent: ask about Avinish
An agent trained on Avinish's profile. Ask about his work, experience, skills, or how to get in touch.
Open to conversations about backend systems, AI agents, and LLM-powered developer tooling.