AI Engineer (R&D) working across the machine learning stack — computer vision, NLP, speech, and LLM fine-tuning — turning research into deployed, production-ready systems.

• Fasih-TTS-V1 — Arabic (Fusha) TTS fine-tuned from Coqui XTTS v2; #1 for intelligibility on the SILMA open-source Arabic TTS benchmark, 1.3% CER (matches human-recording floor). Model · Demo · Write-up
• Whisper-Arabic-finetuning — Quranic ASR fine-tune, tashkeel-aware (CER 0.69% / WER 3.28%) · alignment_quran_recitation — NeMo FastConformer forced alignment for recitation checking
• Quran-Semantic-Retrieval, Hadith_Search, Tafsir_Search — hybrid BM25 + FAISS + verse-anchor semantic search across 7 classical Tafsir books
• Muslim-mode-finetuning — QLoRA persona/behavior fine-tuning (tool routing, scope, measured rulings) · Qwen3-4b-islamic-finetuning
• qwen3-8b-local-server — OpenAI-compatible vLLM server on 4×RTX 2080 Ti

Speech & Audio

LLMs & NLP

Languages & Infra

Serving & MLOps

• FastAPI streaming inference servers · Docker containerization
• Modal (serverless GPU deployment) · ONNX / TensorRT optimization
• Multi-GPU local inference (RTX 2080 Ti ×4, vLLM, Accelerate DDP)

Certifications

• DeepLearning.AI TensorFlow Developer Specialization
• Deep Learning Specialization
• Machine Learning Specialization

Spoken Languages

• AI Instructor, iSchool — teaching Python and applied AI to young students
• Coordinator, AlMentor — organizing AI/tech education events and workshops
• Ambassador, Solve Hub — AI-for-good workshops and events
• Deep Learning Engineer Intern — NLP and computer vision model development & deployment