Available for opportunities

Hi, I'm Shubaib.

Building intelligent systems with Generative & Agentic AI — from RAG pipelines and MCP to autonomous multi-agent workflows. Future full-stack AI Engineer.

scroll

Who I Am

Shubaib
@shubaib-uddin

I'm Shubaib, an aspiring AI Engineer from Pakistan focused on building intelligent systems that solve real-world problems through Generative and Agentic AI.

I work with technologies like RAG, Corrective RAG, and Model Context Protocol (MCP) to create AI systems that retrieve, reason, and respond with accuracy and context awareness.

On the agentic side, I design autonomous AI workflows using LangGraph, LangChain, and LangSmith — building systems capable of planning, decision-making, tool usage, and multi-step reasoning.

My journey is completely self-taught, driven by curiosity, consistency, and building real-world projects from scratch. I enjoy transforming complex AI concepts into practical, scalable applications.

Currently, I'm focused on becoming a full-stack AI Engineer in the new era of Generative & Agentic AI — combining intelligent backends, automation, and modern user experiences into production-ready systems.

What I Work With

AI & LLM Frameworks
LangChainLangGraph LangSmithGenerative AI Agentic AIMulti-Agent Systems CrewAI
RAG & Retrieval
RAGCorrective RAG MCPChromaDB FAISSPostgreSQL
Languages & Tools
PythonJavaScriptPydantic GitGitHubn8n
UI & App Development
JavaScriptHTMLCSS StreamlitTkinter NewsAPIREST APIs
JavaScript Development
ES6+DOM Manipulation Async/AwaitFetch API LocalStorageEvent Handling
Automation
n8n Workflows API Integration Real-time Processing
Prompting & Agentic Workflows
Prompt Design Chain-of-Thought Agentic Systems Tool Use Workflow Orchestration

Things I've Built

01 / AI EVALUATION
AI-Powered Essay Evaluator
A parallel multi-agent LangGraph pipeline where independent AI agents evaluate essays across language quality, critical thinking, clarity, and structure — simultaneously. Final AI model combines all outputs into one comprehensive review with scoring.
LangGraphStreamlit OllamaGroq Pydantic
Watch Demo
02 / AGENTIC AI
Agentic Blog Writing System
Fully autonomous multi-agent system that researches topics via Tavily Search, creates structured blog plans, generates sections in parallel, adds citations, and exports complete Markdown blogs. Pipeline: Router → Research → Planner → Workers → Reducer.
LangGraphOllama TavilyStreamlit
Watch Demo
03 / MCP
AI-Powered Expense Manager
An intelligent expense chatbot that records spending through natural language — no forms, just chat. Auto-logs transactions in Pakistani Rupees, retrieves past records, and delivers category-wise summaries. Powered by MCP + a local LLM for complete privacy and real-time responsiveness.
MCPStreamlit PythonLocal LLM
Watch Demo
04 / CHAT APP
AI Chat Application
A sleek AI chat application with real-time LLM responses and persistent conversation memory. Powered by LangGraph state management with a clean, minimal, fast Streamlit interface focused on smooth user experience.
LangGraphLangChain StreamlitPython
05 / WEB APP
Real-Time News Web App
A lightweight news web app fetching and displaying live news via NewsAPI with a clean, modern Streamlit UI. Practiced real-time API integration, session state management, and image processing with PIL.
PythonStreamlit NewsAPIPIL
Watch Demo
06 / WEB APP
Smart ATM Simulation System
A browser-based ATM simulator with PIN setup, deposit & withdrawal, live balance updates, and transaction history — all persisted via LocalStorage. No backend required.
JavaScriptHTML CSSLocalStorage
Live Demo
07 / AI AUTOMATION
AI Restaurant Assistant
A fully AI-powered restaurant ordering system that handles customer conversations naturally — checking live inventory, processing orders, calculating totals, and managing updates automatically. Powered by OpenAI + n8n automation, with Google Sheets as a real-time backend and WhatsApp API for seamless chat-based ordering.
OpenAIn8n WhatsApp APIGoogle Sheets AI AutomationConversational AI
Watch Demo
08 / ANDROID APP
AI Essay Evaluator — Android App
An Android app that delivers instant, structured essay feedback using Gemini AI. Evaluates across 4 dimensions — Language Quality, Critical Thinking, Clarity, and Structure — generating an overall score, letter grade, criterion-wise feedback, and personalized improvement suggestions. Includes word validation, essay history storage, filterable past evaluations, and a clean dark UI.
GeminiAndroid KotlinAI Evaluation Mobile Dev
Watch Demo
09 / AI CHEF APP
AI Smart Chef (Multi-Stage Recipe Assistant)
A Streamlit-based AI cooking assistant that transforms leftover ingredients into Pakistani recipes. The system uses a two-stage AI agent pipeline: first recommending dishes based on user input or images, then generating full recipe instructions using a specialized instruction agent. Supports image understanding via base64 encoding and text-based ingredient input for hybrid interaction.
Streamlit LangChain Groq LLM Multi-Agent System Image Input (Base64) Python
Watch Demo
10 / MULTI-AGENT SYSTEM
WeddingAI Planner
A parallel multi-agent LangGraph system that plans a full wedding from one request — a Flight agent pulls real routes via the Kiwi MCP server, a Venue agent ranks options with Tavily search, and a Music agent sources tracks and costs from SQLite. All three run simultaneously into one structured Markdown plan.
LangGraph MCP Tavily SQLite Multi-Agent System Groq
▶ WATCH DEMO
11 / AGENTIC AI
AI Speech Generator
A CrewAI application that turns any topic into a polished, ready-to-deliver speech through three specialized agents — Research, Writer, and Review. Splitting the task this way produces noticeably more reliable, well-edited output than a single-pass generation.
The system also includes a final review stage for tone, structure, and audience fit, ensuring every delivery-ready speech matches the intended style and purpose.
CrewAI Multi-Agent System Groq Ollama Python
12 / AI AUTOMATION
Speech Generator — n8n Automation
An end-to-end n8n automation that researches a topic via Tavily and Wikipedia, drafts a structured speech, runs it through a review pass for tone and clarity, and emails the finished result to Gmail.
This workflow separates research, drafting, and review into reusable nodes so every speech stays consistent, polished, and delivery-ready. It also supports rapid updates from new sources without changing the core automation.
n8n AI Agents Tavily Gmail API Workflow Automation
Request Demo

Credentials & Learning

Anthropic Academy
Introduction to Model Context Protocol
TYPE Certificate of Completion
TOPIC MCP · Agentic AI · Tool Use
CREDENTIAL
d23dhm7rdkww
Anthropic Academy
Claude Code 101
TYPE Certificate of Completion
TOPIC Claude Code · AI Agent · Agentic Coding
CREDENTIAL
eucpue5k6owp
LangChain Academy
LangChain Foundations
TYPE Certificate of Completion
TOPIC LangChain · AI Agents · RAG · Prompt Engineering
CREDENTIAL
c39kf5owma
Anthropic Academy
Claude 101
TYPE Certificate of Completion
TOPIC Claude Fundamentals · Prompting · AI Assistants
CREDENTIAL
ocf3ucmgoucp
OpenAI Academy
AI Foundations
TYPE Certificate of Completion
TOPIC AI Fundamentals
CREDENTIAL
in9hj0fbet
IBM SkillsBuild
Introduction to Generative AI
TYPE Certificate of Completion
TOPIC Generative AI Fundamentals
CREDENTIAL
MDL-388
Certificate

Let's Connect

I'm open to opportunities, collaborations, and interesting AI projects. Feel free to reach out — I'd love to connect!