I'm an AI engineer and designer from Saudi Arabia, working across computer vision, voice AI, retrieval-augmented generation (RAG), and agentic systems.
My work spans model evaluation, system architecture, full-stack development, and product design. I focus on building systems that work beyond the prototype, with attention to reliability, usability, and measurable results.
I document my projects, technical decisions, and experiments at turkialshuaibi.com.
HARIS | Computer Vision & Threat Recognition
Led a team of five developing a multi-stage CCTV threat-recognition system combining RT-DETR, BoT-SORT, RTMPose, and ST-GCN. Achieved 0.745 end-to-end F1 on UCF-Crime and reduced per-frame weapon false positives from 8.49% to 0.75%.
Sarjy | Bilingual Voice AI Assistant
Independently built a conversational voice assistant supporting English and Saudi Arabic, with persistent memory, external tool integration, and real-time speech interaction. Built with Next.js, Groq Whisper, and Orpheus.
Onkith | Privacy-Preserving AI
Deputy lead of a six-person team developing a privacy-first voice assistant. Built speech processing components and personally identifiable information (PII) masking capabilities covering 33 entity types.
Madaric | AI Career Intelligence
Independently developed a bilingual career platform combining CV parsing, candidate assessment, opportunity matching, and Saudi job-market insights.
Foreman | Open-Source Agentic Development
Created an eight-stage development playbook for AI coding agents, enforcing requirements gathering, design decisions, implementation checks, and deployment verification.
GitHub Repository · Documentation
- AI & Machine Learning: Computer Vision, RAG, NLP, Speech AI, Model Evaluation, Deep Learning
- Models & Frameworks: PyTorch, RT-DETR, ONNX, Whisper, ST-GCN
- Engineering: Python, TypeScript, Next.js, REST APIs, PostgreSQL, Vector Databases
- Design & Development: Astro, Interactive Web Experiences, AI-Assisted Workflows
AI Engineer | Adalah-AI, VR House
March 2026 to June 2026
Evaluated an AI-powered legal research platform built around Saudi regulatory documents. Tested retrieval quality, citation accuracy, hallucination behavior, API integrations, and database operations. Identified and triaged more than 30 defects.
B.Sc. Artificial Intelligence | Imam Abdulrahman Bin Faisal University
2026 | Second Honors | GPA: 4.51/5.00
AI Specialization | KAUST Academy
June to August 2026
Selected for a highly competitive AI specialization program. Achieved first place in image inpainting and third place in a 100-species image classification challenge.
- Technical Mentor: ByteBank FinTech Hackathon, KFUPM
- Financial Expert: Live 80 Hackathon
- Harvard: CS50's Introduction to Artificial Intelligence with Python
- NVIDIA: Transformer-based Natural Language Processing
- DeepLearning.AI: Mathematics for Machine Learning
- McKinsey.org: Forward Program
