Mohammad Salhab
AI & Machine Learning Engineer · Computer Engineer, 1st in class at BAU
Lebanon · available to relocate immediately · open to remote
+961 81 831 738 · contact@salhabweb.com · linkedin.com/in/mhmdsalhab · salhabweb.com
Summary
Computer Engineering graduate, 1st in class at Beirut Arab University (CGPA 3.99/4.0) and IEEE ICONET26 author. Built a client-churn model that runs in production at Wakilni. Seeking AI / machine-learning engineering roles.
Experience
AI Engineering Intern · Wakilni
Jun – Sep 2026E-commerce logistics & fulfilment · Beirut
- In production: a client churn-risk model built on live operational data, scoring every active client 0–100 (calibrated) with three plain-English reasons (SHAP), plus survival analysis for when clients tend to leave.
- Running as a service, not a notebook: FastAPI backend, React dashboard sorted by revenue at risk, scheduled pipeline, deployed on a hosted server.
- One of two contributors chosen to build the central repository for the programme's approved AI projects.
Technology Intern · Saida Municipality
Jun – Sep 2026Public-sector digital transformation · Saida
- Shipped the Home-screen features of the municipality's citizen-complaints app (Expo / React Native, TypeScript): recent-reports feed, first-time reporting tutorial and in-app tutorial video, in Arabic, English and French.
Founder · SalhabWeb
2024 – PresentWebsites, WhatsApp chatbots and AI agents for businesses
- Build and launch websites, WhatsApp chatbots and AI agents for clients, end to end: from first call to deployment.
Education
B.E. Computer Engineering · AI Track · Beirut Arab University (BAU)
Sep 2022 – Sep 2026- Gamal Abdel Nasser Award (High Distinction): awarded to the top-ranked graduate of each department, university-wide.
- CGPA 3.99 / 4.0 · graduated 1st in Computer Engineering · A or A+ in every graded course but one · final-year project graded A+.
- Coursework: Deep Learning · Machine Learning · Computer Vision · Embedded Systems · Compiler Design · Operating Systems · Databases · Cryptography.
Projects
VocalBlade · AI text-based video editor
Sep 2025 – Jun 2026Final-year project, graded A+
- Edit the transcript and the video edits itself; natural-language requests become word-level cuts through a local LLM (Qwen 2.5 7B). WhisperX, speaker diarization, voice cloning, RIFE/FOMM interpolation. Django + GraphQL, React + TypeScript.
MelodicVision (Melvis) · AI music genre classifier
2026IEEE ICONET26 paper, presented Oct 2026
- 82% test accuracy on GTZAN with only 799 training clips: 30-second clips become 128×128 Mel spectrograms read by a fine-tuned ResNet-18. Grad-CAM explainability; sliding-window ensemble for full songs.
Kimo · voice-controlled robot pet
Spring 2026Embedded Systems lab
- Understands spoken commands in English and Arabic and chases a red object on its own: Whisper + bilingual keyword matching, a local Phi-3 Mini via Ollama, OpenCV tracking; Raspberry Pi 5 brain, Arduino + L298N driving a 4WD chassis. Works offline.
Technical skills
- AI / Machine learning
- PyTorch · scikit-learn · XGBoost · SHAP · survival analysis · MLOps · OpenCV · MediaPipe · Whisper / WhisperX · LLM apps (Ollama, local models)
- Programming
- Python · SQL · C/C++ · JavaScript / TypeScript · Bash · Assembly (8085)
- Backend & web
- FastAPI · Django · Node.js · React · Next.js · REST / GraphQL · MySQL · MongoDB · PostgreSQL
- Systems
- Linux · Git / GitHub · server deployment · Docker · Cloudflare · Raspberry Pi · Arduino
Publication
M. Salhab, Z. Doughan. “Melodic Vision (Melvis): Music Genre Classification via Deep Learning on Mel Spectrograms.” ICONET26, IEEE Computer Society Lebanon Chapter, Oct 2026.
Honors
- Gamal Abdel Nasser Award, High Distinction (Sep 2026)
- Faculty of Engineering top-students list, 4 semesters (1st, 3rd, 2nd, 1st)
- Certificate of Recognition: top of class, first year (2023)
Languages
- Arabic — native
- English — professional working proficiency (IELTS 6.5)