Amine Labibi—
EXPLORERPORTFOLIO
PORTFOLIO-CORE
app
Core Focus
Enterprise RAGDocling + Qdrant
AI AgentsLangGraph (HITL)
Full-StackReact + FastAPI
developer_info

Casablanca, Morocco

amineib135@gmail.com

0777411375

portfolioapppage.tsx
TypeScript React

/**

* @author Amine Labibi

* @role AI Full-Stack Developer | React • Node.js • Python • Generative AI

* @specialization Enterprise RAG, LangGraph Workflows, High-Precision Search

* @location Casablanca, Morocco | Available for opportunities

*/

Production AI Engineering & Full-Stack Systems

Amine Labibi

AI Full-Stack Developer|React • Node.js • Python • Generative AI

AI Full-Stack Developer with hands-on experience building end-to-end web applications and Generative AI solutions. Skilled in JavaScript, React.js, Node.js, Express.js, Python, and FastAPI. Experienced in developing REST APIs, integrating LLMs, building RAG pipelines, and implementing document processing and intelligent question-answering workflows. Combines full-stack development capabilities with practical AI engineering and an Economics and Management background.

LOCATION
Casablanca, Morocco
SPECIALTY
Enterprise RAG & Agents
EDUCATION
Economics & Management
STATUS
Open to Opportunities
// 01. Professional Overview & Philosophy

export const developerProfile = () => {

---
name: "Amine Labibi"
role: "AI Full-Stack Developer"
location: "Casablanca, Morocco"
degree: "Licence in Economics & Management (Hassan II University)"
core_focus: ["Enterprise RAG", "LangGraph Agents", "Docling", "Qdrant", "FastAPI", "React"]
---

I am an AI Full-Stack Developer specializing in creating resilient, real-world web applications and autonomous Generative AI pipelines. Based in Casablanca, Morocco, my focus sits at the intersection of modern full-stack development and production-grade AI engineering.

Enterprise RAG & Search

Layout-aware parsing with IBM Docling, hybrid dense + BM25 search in Qdrant, ChromaDB, LlamaIndex, and low-latency inference using NVIDIA NIM.

Agentic Workflows

Orchestrating reliable multi-step agents with LangGraph, confidence scoring, automated response verification, and Human-in-the-Loop review routing.

Economics & Analysis Edge

Economics & Management degree gives a deep grasp of financial statements, balance sheets, unit metrics, and business logic that generic developers lack.

Whether developing full-stack architectures (React, Node.js, Express, MongoDB, PostgreSQL) or designing decoupled Python FastAPI AI microservices with Groq and OpenRouter, I build clean, testable, and audited systems with zero tolerance for arithmetic hallucinations or flaky retrieval.

}

// 02. Featured Deployments & Engineering Implementations

const selectedProjects = [

Enterprise RAG

EnterpriseRAG Knowledge Base

An enterprise-grade Retrieval-Augmented Generation (RAG) platform specialized in technical documentation, engineering manuals, and dense tabular data.

•
Architecture & Parsing: Built with IBM Docling for layout-aware document and tabular structure extraction.
•
Hybrid Search Engine: Implemented Qdrant Hybrid Search combining dense neural semantic embeddings with sparse BM25 lexical search for high precision on domain-specific engineering jargon.
•
High-Performance Inference: Powered by NVIDIA NIM Models for optimized enterprise inference, grounded answer generation, and low-latency retrieval synthesis.
IBM DoclingQdrantNVIDIA NIMPythonHybrid Search (Dense + BM25)FastAPIRetrieval Eval
Financial RAG

Financial Statement RAG

Specialized financial analysis platform combining unstructured disclosures with structured financial accounting records and deterministic validation.

•
Hybrid Retrieval Architecture: Built a financial analysis platform combining semantic vector search (ChromaDB + LlamaIndex) for MD&A and footnotes with normalized relational storage (PostgreSQL) for structured financial facts.
•
Deterministic Calculations: Implemented exact mathematical calculations (YoY growth, liquidity, leverage, and margins) using Python decimal to eliminate LLM arithmetic hallucinations.
•
Grounding & Auditability: Designed a verification layer that reconciles LLM statements against tabular database records and validates citation sources with strict unit/scale tracking.
PythonChromaDBLlamaIndexPostgreSQLPython DecimalVerification LayerAuditability
AI Agent

Human-in-the-Loop AI Agent (LangGraph)

Autonomous agent system designed for high-consequence enterprise workflows with automated confidence scoring and human oversight.

•
Workflow Orchestration: Designed an agent workflow using confidence scoring and human review routing for uncertain cases.
•
Evaluation & Reliability: Included automated response verification and an interactive feedback loop to support reliable outputs.
PythonLangGraphLangChainReAct AgentsConfidence RoutingHuman-in-the-LoopEvals
Full-Stack AI

ShopAI – AI-Powered Shopping Platform

Full-stack AI-powered e-commerce platform with product browsing and an interactive conversational shopping assistant.

•
Conversational Shopping: Integrated a Python FastAPI AI service with Groq LLMs for conversational assistance and structured, product-aware recommendations.
•
Backend & Data: Built REST APIs using Node.js and Express.js, with MongoDB for product and application data management.
•
Decoupled Architecture: Designed a clean separated frontend, backend, and AI-service microservices architecture.
React.jsJavaScriptNode.jsExpress.jsMongoDBMongoosePythonFastAPIGroq API
Full-Stack AI

AI Blog – Content Generation Platform

Full-stack AI-powered blogging platform with an interactive content creation studio and automated writing assistance.

•
LLM Writing Features: Integrated LLM-powered features for blog generation, content improvement, grammar correction, title generation, and category suggestions.
•
Dedicated Microservices: Built backend APIs with Express.js and a dedicated Python FastAPI service for high-performance AI processing.
•
Data & Security: Implemented application functionality using MongoDB, authentication-related dependencies, and REST API communication.
React.jsJavaScriptNode.jsExpress.jsMongoDBPythonFastAPIGroq APIREST APIs

]

// 03. Comprehensive Technical Skill Matrix

interface AmineTechnicalSkills {

Generative AI & Agents
LLM Integration (Groq, OpenRouter)
LangChain & LangGraph
ReAct Multi-Agent Workflows
NVIDIA NIM High-Performance Models
Prompt Engineering & Structured Outputs
Human-in-the-Loop Review Guardrails
RAG & Vector Search
IBM Docling (Layout-Aware Tables)
Qdrant Hybrid Search (Dense + BM25)
ChromaDB Vector Store
LlamaIndex Framework
Embeddings & Retrieval Evaluation
Deterministic Verification Layers
Backend & Microservices
Python & FastAPI
Node.js & Express.js
RESTful API Architecture
JWT Authentication
HTTP-only Cookies Security
Microservices Communication
Frontend Engineering
React.js (Hooks, State, Context)
JavaScript (ES6+) & TypeScript
Tailwind CSS & Responsive Design
Next.js App Router
Component-Driven Architecture
Interactive Chat & AI Interfaces
Databases & Persistence
PostgreSQL (Relational Facts)
MongoDB & Mongoose
Supabase (Auth & Postgres)
Vector Indexing (HNSW, Cosine)
Schema Normalization
Audit Trails & Fact Tables
Tools & Workflow
Git & GitHub Workflows
Postman API Testing
Vercel Deployment
Clean Architecture & Modular Code
Deterministic Decimal Arithmetic
Linux & PowerShell CLI

}

// 04. Education, Training & Certifications

const credentials = {

Academic Education

Bachelor's Degree (Licence) in Economics & Management

Hassan II University • 2025–2026

Rigorous analytical training in managerial economics, corporate finance, market structures, and quantitative decision-making. Directly empowers domain-aware AI engineering for financial statement RAG, business automation, and enterprise data reasoning.

LANGUAGES
ArabicAdvanced
EnglishIndependent
FrenchElementary
Training & Professional Certifications
Geeks Institute
Web Development, AI Agents & RAG Training
Intensive hands-on mastery in building end-to-end full-stack applications, LangGraph agents, and production RAG.
Coursera & DeepLearning.AI
Build RAG Applications: Get Started
Vector Databases for RAG: An Introduction
Generative AI Engineering with LLMs (modules completed)
ALX
Professional Skills & Software Development Fundamentals

}

// 05. Environment Variables & Connection Endpoint

export connection = connect()

I am actively open to discussing full-time roles, AI consulting, or high-impact engineering projects. Feel free to copy connection variables or send a direct dispatch below.

.env.production
DEVELOPER_NAME="Amine Labibi"
ROLE="AI Full-Stack Developer"
LOCATION="Casablanca, Morocco"
EMAIL="amineib135@gmail.com"
PHONE="0777411375"
GITHUB="https://github.com/Amine4144244"
PORTFOLIO="https://portfolio-silk-one-40.vercel.app"
DEGREE="Economics & Management"
execute_dispatch()

amine@casablanca:~/enterprise-rag$ python -m docling.extract --doc "manual_v4.pdf" --tables

[INFO] Docling layout analysis complete: 34 pages parsed. 12 dense tables extracted with structural markdown fidelity.

amine@casablanca:~/enterprise-rag$ qdrant-hybrid --query "turbine cooling specs" --top-k 5

[HYBRID SEARCH] Dense vector (cosine) + BM25 sparse lexical fused via Reciprocal Rank Fusion (RRF). Accuracy: 99.1%.

amine@casablanca:~/portfolio$ █

Portfolio*
TypeScript React
VS Code Dark+