Danah Alakrash.
AI/ML Engineer & Full-Stack Developer — Computer Vision, NLP & Scalable Web Systems
About Me
I'm an engineer who cares just as much about the people using the software as the code behind it. I build computer vision and NLP systems—from real-time detection pipelines to semantic search engines—and pair them with full-stack web interfaces that make the underlying AI genuinely usable. For me, the best part of this job isn't just solving a hard problem in isolation—it's taking a massive, confusing challenge and figuring out how to make it simple, intuitive, and genuinely helpful for someone else.
I don't build in a silo. I thrive on sitting down with a team, breaking down a messy problem, and trying things out until it clicks. I've shipped models to edge devices and to the browser, building the backends and interfaces around them. If I have a superpower, it's communication—turning a complex idea into something anyone in the room can follow. I want to build great things alongside people who push each other to grow.
Open to AI/ML Engineer and Full-Stack Developer roles — graduated June 2026.
Background & Experience
Qassim University
Bachelor of Science in Computer Science • 06/2026
Graduated with a 4.37/5 GPA and Second Class Honor. Along the way, I learned to design efficient algorithms, manage complex data structures, and approach software development with an analytical mindset — skills I now apply to build robust, real-world systems.
AI & Full-Stack Developer Intern
Nighat AI & IoT • 06/2025 - 08/2025
Built a two-stage vision system for industrial date sorting — YOLOv8 detection into grade classification, 98%+ accuracy on a held-out test set. Deployed to Raspberry Pi at 10+ FPS under production lighting.
Featured Projects
LevelUp
Bilingual (Arabic/English) AI interview coach — multilingual SBERT CV-to-job matching plus a Gemini 2.5 Flash coach that generates questions from the candidate's own CV. Gaze and posture analysis run client-side via MediaPipe, so no video leaves the browser.
Touchless Air Mouse
Real-time hand-gesture mouse control using 21-point 3D landmark tracking with custom jitter filtering for stable, hardware-free navigation. Runs in real time on CPU — no GPU required.
Watch Price Predictor
Random Forest pipeline trained on 45,000+ luxury watch listings — R² of 0.73 and $6,683 MAE, outperforming a linear baseline.
Semantic Text Matcher
In-browser NLP engine using SBERT (MiniLM-L6-v2) to compute 384-dim semantic similarity between texts — runs entirely client-side via Transformers.js.
Salary Predictor
Full-stack salary prediction model (0.93 R²) served through a Next.js API route for real-time inference with zero cold-starts.
AI Color Architect
K-Means clustering pipeline that extracts and ranks dominant color palettes from any uploaded image.
A* Visualizer
Interactive A* search visualizer comparing heuristic performance — demonstrates Manhattan Distance's dominance over Misplaced Tiles across test configurations.
Skills & Certifications
Core Stack
AI & Machine Learning
Deep Learning, YOLO, NLP, Fine-Tuning. Click for details.
Development & DevOps
React.js, FastAPI, Streamlit, Docker. Click for details.
IBM Certifications
4x Credentials in LLMs, RAG, and Fine-Tuning. Click to view.
LLMs & Generative AI
RAG, LangChain, Fine-Tuning, Agents. Click for details.