Atindra Mishra
AI Engineer &
Systems Builder
Agentic AI · LLMOps & Evals · Production RAG · Full-Stack
I build AI systems that ship: multi-agent orchestration with LangGraph, eval-driven RAG pipelines, and LLMOps infrastructure with observability from day one.
3+
Projects Deployed
2+
Years in Software Dev
1+
Years in AI & Gen AI
20+
AI & Web Frameworks
// featured stack
scroll to explore
// about
About Me
I'm an AI Engineer focused on building systems that are production-ready from the first commit. My work spans multi-agent orchestration with LangGraph, eval-driven RAG pipelines, and the LLMOps infrastructure that keeps AI applications reliable at scale.
At Invasion Media I lead AI feature development architecting LangGraph agent pipelines, wiring LangSmith observability, and running systematic evals before any model change ships.
I believe the gap between an “AI demo” and an “AI product” is eval coverage and observability. That's where I spend my energy.
// skills
Core Competencies
Depth where it matters, breadth across the stack
Agentic AI & Orchestration
ExpertLangGraph, LangChain, AI agents, tool use, MCP, production RAG pipelines
LLM & RAG Systems
ExpertHybrid BM25+vector retrieval, ChromaDB, Pinecone, pgvector, RAGAS evals
AI Evaluation & LLMOps
ProficientDeepEval, RAGAS, Langfuse, CI-gated eval pipelines, cost tracking
Python & FastAPI
ExpertAsync APIs, JWT auth, WebSocket, production backend systems
Full-Stack (React/Next.js)
ExpertReact.js, Next.js, TypeScript, Tailwind CSS, Zustand
Full Technology Stack
Agentic AI & Orchestration
AI Evaluation & Observability
LLMs & Model Infrastructure
Backend & APIs
Frontend & Full-Stack
Databases & DevOps
// projects
Projects
Concierge Support ADK
AI Projects
Portfolio-ready AI support concierge demo with a FastAPI backend and a Next.js frontend. Demonstrates a persistent support assistant built with Google ADK — creates support sessions, routes chat turns through specialist agents, answers documentation questions from a local RAG corpus, persists state, and exposes structured traces in the UI.
Stack
Key Features
- →Google ADK routing between knowledge and account specialist agents
- →Local RAG over bundled Markdown docs with Chroma vector store
- →Structured trace viewer with route, latency, tool calls, and retrieved chunks
- →Session creation with user ID and plan tier, with full state persistence
AI Ad Analyzer
AI Projects
Lightweight app for competitor ad intelligence. Creates a brand profile from a real website, fetches competitor ads from Meta Ad Library through Apify, analyzes copy and images with Groq models, persists structured intelligence in SQLite, and powers a grounded creative strategy chat.
Stack
Key Features
- →Scrapes brand websites and extracts positioning, tone, audience, and value propositions
- →Fetches real Meta Ad Library creatives and analyzes copy hooks, CTAs, and messaging angles
- →Groq llama-4-scout vision analysis: style, UGC vs produced, product visibility
- →Grounded creative strategy chat over brand profile and analyzed competitor ads
Company Intelligence Agent
AI Projects
A LangGraph-powered company research agent that generates structured intelligence reports from a company name, website URL, or LinkedIn link. Researches public web signals — funding, leadership, competitors, news, hiring trends, and tech stack — then outputs a clean markdown report with sources and an agent execution trace.
Stack
Key Features
- →Accepts company name, website URL, or LinkedIn company link as input
- →Researches funding, leadership, competitors, news, hiring trends, and tech stack signals
- →Generates structured markdown intelligence report with confidence assessment and source list
- →Exposes full LangGraph agent execution trace alongside the report
// experience
Work History
From freelance web development to production AI systems
Gen AI & Full Stack Developer
Aug 2025 – PresentInvasion Media
Built and deployed production RAG pipelines using LangChain, OpenAI API, and open-source LLMs (Ollama, Groq) with vector databases (Pinecone, ChromaDB, pgvector) serving international clients with large-scale semantic search. Engineered AI agent workflows for Discord automation and email processing pipelines, integrating LLM routing between cloud and local models to optimize inference cost. Deployed and monitored applications on AWS EC2/S3, Vercel, and Render with CI/CD pipelines and observability instrumentation.
Team Size
International clients
Projects
RAG + agent pipelines
Impact
Reduced manual processing time
Full Stack Blockchain Developer Intern
Feb 2025 – Jul 2025Cluster Protocol
Built and deployed 3+ decentralized applications (DApps) with MetaMask and WalletConnect integrations using Solidity, Ethers.js, and Web3.js. Developed end-to-end fullstack DApps using MERN stack and Next.js, handling UI/UX through to smart contract integration.
Team Size
Remote team
Projects
3+ DApps deployed
Impact
MetaMask & WalletConnect
Freelance Full Stack Developer
Apr 2024 – Dec 2024Real Space Group & RSVP Group
Delivered 2 responsive real estate web platforms using React.js and Node.js with RESTful APIs, mobile-first design, and property listing management.
Team Size
Freelance
Projects
2 platforms delivered
Impact
Mobile-first, property listing
// contact
Ready to Connect?
Open to AI Engineer roles, agentic systems collaborations, and discussions about LLMOps, evals, and production AI architecture.
// or reach me on