m
mohamedh02

Mohammed H

@mohamedh02

Data Scientist and AI Engineer

Territorios Palestinos
Inglés, Árabe
Parte de la información aparece en idioma inglés.
Sobre mí
AI Engineering Intern at Kawkab, an applied AI firm serving MENA enterprises, where I build bilingual Arabic/English LLM products with Claude, OpenAI and Gemini APIs, FastAPI and React. Data Scientist with a Computer Engineering degree and a 660-hour ML bootcamp. I build ML models (churn prediction, PR-AUC 0.873), Arabic NLP (sentiment classifier, macro-F1 0.914), LLM apps and chatbots with guardrails, and Python data analysis and forecasting. Native Arabic, fluent English. Clear communication and clean, documented code.... Lee más

Habilidades

m
mohamedh02
Mohammed H
desconectado • 

Revisa mis servicios

Visión Artificial
I will build a machine learning model for your data in python
DeUSD 90 / proyectoMás información
Procesamiento de lenguaje natural
I will do arabic and english nlp and sentiment analysis in python
DeUSD 90 / proyectoMás información

Experiencia laboral

Kawkab_AI

AI Engineering Intern

Kawkab AI • Tiempo parcial

Jul 2026 - Present • 3 mos

Part-time AI engineering internship at Kawkab, an applied AI firm serving MENA enterprises. I build bilingual (Arabic/English) LLM products end to end. - Kawkab Content Studio: bilingual marketing-copy generator in Next.js 15, React 19 and TypeScript, with validated API routes, a reusable brand-kit system, bulk generation of up to 10 variants, and a 15+ component design system. - Brand-compliance scoring endpoint that rates generated copy 0-100 against a brand's tone, values and approved facts, and returns itemized issues instead of pass/fail. - Bilingual content-generation MVP on FastAPI, SQLAlchemy and the Anthropic Claude API: tenant-scoped data model, approved-fact grounding, constraint validation with bounded auto-retry, and a mandatory human review queue, tested with pytest. - Client Scoping Assistant backend (TypeScript/Express, SQLite) in a 5-person team: gap analysis over a structured brief, question prioritization, multi-turn session state, and an API contract agreed up front so the frontend was built in parallel during a 1.5-day sprint. - AI student-support platform (fellowship final project, 2-person team): FastAPI/PostgreSQL backend, React frontend, Gemini assistant and OpenAI-based evaluator, Docker deployment.