Everything, everywhere

My work, at a glance

Anuj

AI Architect · M4ST

Available for work

📍 Bareilly, India

Tech Stack 21 tools

⚡LangGraph

🤝CrewAI

🧠ChromaDB

🔗MCP Protocol

🐍Python

🎯

The Mission

Build autonomous AI systems that genuinely work without human babysitting — local-first, zero cloud cost, production-grade. 21 MCP servers across 11 free LLM providers. This is not a demo. This is production.

MCP Servers

LLM APIs

Monthly Cost

∞

Scale

⚡ FLAGSHIP PROJECT

M4ST Agent Orchestration Stack

The pinnacle of local-first AI architecture. A complete multi-agent orchestration system running on LangGraph and CrewAI, coordinating specialist agents across 21 MCP servers. ChromaDB-backed semantic memory enables agents to remember across sessions. 11 free LLM providers routed intelligently via LiteLLM. Zero cloud cost. Production ready. Runs completely offline.

LangGraphCrewAI ChromaDBMCP OllamaPython LiteLLM

▶Orchestrator Agent

├→Research Agent· web MCP

├→Code Agent· exec MCP

├→Memory Agent· chroma

└→File Agent· fs MCP

→LLM: Gemini / Groq / Mistral (free)

→Memory: ChromaDB local vector store

21MCP Servers

11LLM Providers

$0Monthly Cost

∞Scale

🔍

OSINT Intelligence System ↗

Autonomous intelligence gathering. Multi-source fusion, anomaly detection, pattern recognition, automated reports. Zero human intervention.

LangChainOSINT APIsAgents

🧠

Semantic Memory Engine ↗

Persistent vector memory with semantic clustering, relevance decay, and context-aware injection. Agents that genuinely remember.

ChromaDBEmbeddingsRAG

🐙

Smart GitHub Manager ↗

AI repo intelligence. Auto docs, PR analysis, issue triage, code review agents. Live on GitHub right now.

GitHub APIMCPAI Agents

📍

Bareilly

Uttar Pradesh, India
Working globally, remotely

IST UTC+5:30

🐙

m4stanuj

Open-source agent systems
MCP servers & AI tools

View Profile ↗

Let's build
something great

✉ m4stanuj@gmail.com

Services

What I offer

⚡Multi-Agent System Design→

🔗MCP Server Development→

🤖AI Pipeline Automation→

💡LLM Orchestration & Routing→

🧠Semantic Memory Systems→

🔍OSINT & Intelligence Systems→

"Intelligence should be autonomous, not dependent — local, not leased. The edge is the future."

Anuj · M4ST · 2026

Connect

Find me online

🐙GitHub↗ 💼LinkedIn↗ 𝕏X / Twitter↗ 🌐Portfolio↗

About

Building AI
differently

I'm Anuj — a solo AI systems engineer from Bareilly, Uttar Pradesh, India. I specialize in building autonomous multi-agent AI systems that think independently, collaborate across specialized agents, and execute complex workflows without human oversight.

My entire stack is local-first: 21 MCP servers providing rich tool access, intelligent routing across 11 free LLM providers, ChromaDB for persistent semantic memory, and CrewAI + LangGraph for orchestration. The whole system runs in production at exactly $0 per month.

I'm available for AI freelance engagements worldwide — agent architecture, MCP development, pipeline automation, and AI infrastructure consulting.

🏠

Local-First Architecture

Everything runs on your hardware. No cloud dependency. No data leaving your perimeter. No monthly SaaS bills. Offline capable by design.

💰

Zero-Cost Operations

Production AI at $0/month. 11 free LLM providers routed intelligently. Enterprise capability at zero recurring cost — not a demo, actual production.

🤖

Genuine Autonomy

Agents that decompose tasks, collaborate, build semantic memory, and execute — without asking for permission. Not wrappers. Real autonomous systems.

🔧

Composable by Design

21 MCP servers, each purpose-built. Swap, extend, compose infinitely. Every component is a modular building block for any AI workflow.