// aiprov.xyz — initialized
00:00:00 --- --- -- ----
Welcome to DURBEK'S AI Laboratory

AI LABORATORY

Researching, building and deploying intelligent agent systems. From zero to production — experimenting with the future of AI.

View Projects Meet Agents
0
Projects Built
0
Agents Deployed
0
Experiments Run
∞
Ideas Queued
aiprov@lab ~ terminal
aiprov@lab:~$ ./initialize_lab.sh
Lab environment initialized
AI agents loaded: n8n + CrewAI + LangGraph
LLM connected: Groq (Llama 3.3) | Gemini Flash
→ aiprov.xyz ready for deployment
aiprov@lab:~$
Research & Projects

Lab Experiments

Multi-Agent
AGENT ORCHESTRATOR
Multiple AI agents working together — research, write, deploy. Fully automated pipeline using CrewAI framework.
Building→
Automation
TELEGRAM → SITE BOT
Send a message on Telegram — agent publishes it to this site. n8n workflow + GitHub API integration.
Building→
Research
LLM COMPARISON LAB
Testing Groq, Gemini, Ollama side-by-side. Speed, accuracy, cost analysis across different tasks.
Planned→
Tool Use
WEB RESEARCH AGENT
Agent that autonomously browses the web, summarizes content, and produces structured reports.
Planned→
Vision
IMAGE ANALYSIS AGENT
Upload images — agent describes, classifies, and extracts data. Built with multimodal LLMs.
Planned→
Coming Soon
NEXT EXPERIMENT
New experiment loading... The lab is always running.
Queued+
Agent System

Active Agents

🧠
ORCHESTRATOR
Master agent that delegates tasks to other agents and monitors results.
LangGraph
🔍
RESEARCHER
Searches the web, reads articles, and compiles structured knowledge.
Groq + Search
⚙️
SITE MANAGER
Controls this website — posts, updates content, changes layouts on command.
n8n + GitHub
💬
COMMUNICATOR
Handles Telegram messages, turns commands into lab actions instantly.
Telegram Bot
Tech Arsenal

Tools & Stack

Every tool here is battle-tested in real lab experiments. No fluff — only what actually works.

Orchestration
⚙️
n8n
Workflow Automation
Visual workflow builder. Connects APIs, triggers agents, automates everything — from Telegram to GitHub.
Self-hostedFree
🧠
LangGraph
Agent State Machine
Graph-based agent orchestration. Builds complex multi-step reasoning loops with full state management.
PythonOpen-source
🤝
CrewAI
Multi-Agent Framework
Role-based agent crews. Researcher + Writer + Deployer all working together on one mission.
PythonOpen-source
🌊
Flowise
Visual LLM Builder
Drag-and-drop RAG pipelines. Build chatbots and agent flows visually — no code required.
No-codeFree
LLM Inference
⚡
Groq
Ultra-fast Inference
LPU-powered inference. Llama 3.3 at insane speed — 300+ tokens/sec. Free tier is generous enough for real projects.
Llama 3.3Free tier
✨
Gemini Flash
Google AI
1M token context window. Multimodal — text, image, video. Free tier with generous daily limits via Google AI Studio.
MultimodalFree tier
🐒
Ollama
Local Models
Run any open-source model locally. Zero API costs, full privacy. Llama, Mistral, Qwen — all on your machine.
100% LocalOffline
Deploy & Language
🚀
Vercel
Deployment
Push to GitHub → live in 30 seconds. Free tier covers everything for a personal lab. Edge functions included.
CI/CDFree
🐍
Python
Primary Language
The backbone of all agent logic. LangChain, CrewAI, LangGraph — everything runs on Python in this lab.
3.11+AI/ML
🐘
GitHub
Version Control + API
Not just version control — the GitHub API lets agents commit code, update files, and deploy automatically.
REST APIFree
About the Lab

Who's Running This

This is AIPROV LAB — a personal AI research laboratory where I build, test, and experiment with intelligent agent systems.

My goal: deeply understand AI agents and deploy them in real-world applications. Every project here is a live experiment.

Tools of choice: n8n, CrewAI, LangGraph, Flowise for orchestration. Groq and Gemini for inference — both free-tier.

Everything is open, documented, and reproducible. The lab never stops running.

  • 🤖AI Agents
  • ⚙️n8n / Flowise
  • 🐍Python / LLM
  • 🌐Web / Deploy
  • 🔬Research
Track 01
AIPROV LAB