AI Engineer focused on LLM applications, applied machine learning, Cloud AI, and production-minded AI systems.
Website | LinkedIn | GitHub | Email
I build AI projects that connect real users, domain context, and deployable software. My strongest work sits at the intersection of Arabic-first AI experiences, context-aware assistants, intelligent search, security analytics, computer vision, and cloud architecture.
| Area | What it means in my work |
|---|---|
| LLM applications | Local and cloud-oriented assistants that use domain context instead of generic chatbot behavior. |
| Applied machine learning | Prototypes for anomaly-style detection, computer vision, behavioral analytics, and decision support. |
| Cloud AI | Systems designed around APIs, deployment paths, storage, observability, and managed AI services. |
| Arabic AI products | Interfaces and knowledge flows built for Arabic-first users and bilingual product contexts. |
| Engineering communication | READMEs, architecture notes, tradeoffs, and limitations that help reviewers understand the system quickly. |
| Project | Engineering focus | Why it matters |
|---|---|---|
| Abdulelah AI Portfolio | Next.js, TypeScript, structured content, SEO, portfolio assistant | Connects the website, project case studies, role-specific resumes, and AI assistant into one portfolio system. |
| ChatUB | Arabic LLM assistant, semantic search, local AI, Flask, Ollama | Demonstrates how a university-specific assistant can answer student questions using trusted academic content. |
| Absher Insight AI | AI security analytics, FastAPI, synthetic behavior data, anomaly-style rules | Explores proactive risk detection for government-style digital services without exposing real user data. |
| Architect of Intelligence | TanStack Start, TypeScript, structured AI workflows, Zod validation | Shows modern AI platform architecture with server-side validation, bilingual output, and fallback behavior. |
| Stadium Gate Monitor | YOLO, OpenCV, Flask APIs, operations dashboard | Applies computer vision to crowd monitoring, gate status, alerting, and staff-distribution decisions. |
| Alpha AI Innovations | Arabic AI advisory concept, React, TanStack Router, structured service catalog | Shows Arabic-first AI product thinking with conservative prototype boundaries and documented validation gaps. |
I prefer AI systems that are understandable, constrained, and useful:
- Ground responses in real domain material whenever possible.
- Separate prototypes from production claims.
- Design data boundaries before model behavior.
- Make APIs, fallbacks, and failure modes explicit.
- Document architecture, setup, limitations, and future work.
- Avoid invented metrics, inflated achievements, or vague AI language.
| Layer | Tools and concepts |
|---|---|
| AI and data | Python, NLP, LLM applications, semantic search, SentenceTransformers, anomaly detection, computer vision |
| Backend | FastAPI, Flask, API route design, Pydantic, Zod, REST-style services |
| Frontend | TypeScript, React, Next.js, TanStack Start, Vite, Tailwind CSS |
| Cloud and deployment | Google Cloud exposure, Azure AI Services exposure, Vercel, Cloud Run concepts, Cloud Storage concepts, BigQuery concepts |
| Product delivery | Arabic-first UX, technical writing, architecture diagrams, dashboard design, privacy-aware flows |
| Category | Repositories |
|---|---|
| Flagship AI projects | ChatUB, absher-insight, architect-of-intelligence, Stadium |
| Portfolio system | Abdulelah |
| Hackathon and product concepts | alpha-ai-innovations, midad-landing |
| Archived learning projects | Older static exercises, placeholders, duplicates, and incomplete scaffolds are intentionally not highlighted. |
Detailed ranking and curation notes: Repository Strategy
Recruiter review sequence: Recruiter Review Path
If you are evaluating my work for an AI engineering role, start here:
- ChatUB for LLM and Arabic NLP work.
- absher-insight for AI security and decision-support thinking.
- architect-of-intelligence for TypeScript AI platform architecture.
- Stadium for computer vision and operational dashboards.
- Abdulelah for the full portfolio system and technical presentation.
- alpha-ai-innovations for Arabic AI product and advisory-service concept work.
I keep public claims conservative. Projects are documented with:
- What the system does
- What architecture it uses
- What is prototype-level
- What needs more evaluation
- What should be improved before production use
Internal or employer-owned work is not used as public portfolio evidence.
- Website: abdulelah.de
- LinkedIn: linkedin.com/in/abdulelah-alkhathami-853845311
- GitHub: github.com/Abdulel3h
- Email: me@abdulelah.de