Open to opportunities

Doaa Mostafa

@doaamostafa1

Message

I build end-to-end AI products with agentic workflows, RAG systems, and full-stack web applications.

Egypt

Message

What I'm looking for

I'm looking to build end-to-end AI-powered products where I can develop agentic workflows, RAG systems, APIs, and responsive user interfaces while growing through hands-on product delivery.

I've built production-grade AI applications at Next Technology, including a LangGraph customer-support platform with ChromaDB RAG retrieval and a real-time Next.js interface powered by Llama 3. I also created a seven-agent content pipeline for research, SEO generation, adaptation, evaluation, and asynchronous publishing.

Across internships at Nile University, Excelerate, and the National Bank of Egypt, I've delivered React applications, churn models, and Arabic RAG systems. My work spans LLM APIs, REST APIs, FastAPI, Python, React, Next.js, TypeScript, AWS, and Docker—from concept through deployment.

Experience

Work history, roles, and key accomplishments

Frontend Development Intern

Nile University

Apr 2026 - Jun 2026 (2 months)

Developed a TypeScript recipe-blog SPA with search, recipe details, and cooking-tips pages, using Zustand for state management and TanStack Query for data caching. Built a reusable component library with shadcn/ui and Radix primitives on Tailwind CSS v4.

React TypeScript Tailwind Shadcn Ui React Router Zustand Radix Tanstack Query

Agentic AI Intern

Next Technology

Mar 2026 - Jun 2026 (3 months)

Developed a production-grade LangGraph-based customer support platform with RAG retrieval and a real-time Next.js interface. Built a 7-agent LangGraph pipeline for content creation and an agentic AI shopping assistant using CrewAI.

LangGraph GROQ ChromaDB RAG Next.js Celery fastAPI TypeScript Crew AI

AI Data Analyst Insights Intern

Excelerate

Dec 2024 - Jan 2025 (1 month)

Cleaned and preprocessed raw datasets and performed feature engineering for churn modeling. Trained an XGBoost Classifier achieving an ROC-AUC of 0.949, with hyperparameter tuning yielding an ROC-AUC of 0.935 and 91.1% recall.

Python Pandas EDA Feature Engineering XGBoost Hyperparameter Tuning

NLP Engineer Intern

National Bank of Egypt

Jul 2024 - Oct 2024 (3 months)

Developed a Retrieval-Augmented Generation (RAG) system for Arabic response generation, fine-tuned on 1,000+ NBE data entries. Designed and implemented 3 web-scraping pipelines and integrated 2 LLMs for constraint-driven Arabic responses, improving response precision by 30%.

BeautifulSoup Selenium Python GROQ Flask Hugging Face Embeddings