Available for Opportunities

AI ENGINEER
& FULL-STACK
DEVELOPER

20 years of engineering experience. Building LLM pipelines, RAG systems, and scalable SaaS platforms.

View Projects Get in Touch rlyxa.co.uk
LLM Fine-Tuning RAG Systems React & Next.js LangChain Azure ML SaaS Architecture Regulatory Compliance Full-Stack Development LLM Fine-Tuning RAG Systems React & Next.js LangChain Azure ML SaaS Architecture Regulatory Compliance Full-Stack Development
About

Hey! I'm Mohammad.
Building the future with AI.

Highly accomplished AI Engineer and Full-Stack Developer with 20 years of progressive software engineering experience and a specialized focus on Artificial Intelligence and Machine Learning since 2023.

Proven expertise in architecting, developing, and deploying cutting-edge AI systems, including LLM pipelines, RAG systems, and fine-tuned models. Adept at leading end-to-end product development, building scalable SaaS platforms, and driving innovation with a deep understanding of responsible AI, regulatory compliance (GDPR, AML/CFT), and cross-domain product thinking.

0
Years of Experience
0
Synthetic Training Pairs Generated
0
AI-Powered SaaS Tools Built
Expertise

Technical Skills

AI
AI / ML / Deep Learning
Hover to see skills
AI AI / ML / Deep Learning
  • LLM Fine-Tuning (LoRA/QLoRA)
  • RAG Pipelines
  • LangChain
  • OpenAI API
  • Claude API
  • Gemini
  • FAISS
  • Pinecone
  • Prompt Engineering
  • PyTorch
  • TensorFlow
  • Scikit-learn
  • XGBoost
  • MLflow
  • SHAP
  • NLP
FS
Full-Stack Development
Hover to see skills
FS Full-Stack Development
  • React
  • Next.js
  • Node.js
  • TypeScript
  • Remotion
  • Tailwind CSS
  • HTML/CSS
  • REST APIs
  • SSE
  • WebSockets
CL
Cloud & DevOps
Hover to see skills
CL Cloud & DevOps
  • Azure ML
  • Azure Data Factory
  • Azure Blob
  • Docker
  • GitHub Actions
  • CI/CD
  • Vercel
DB
Data & Databases
Hover to see skills
DB Data & Databases
  • SQL
  • PostgreSQL
  • MongoDB
  • Firebase
  • Vector Stores
  • JSONL Pipelines
CO
Compliance & AI Safety
Hover to see skills
CO Compliance & AI Safety
  • GDPR
  • AML/CFT
  • FATF
  • NHS Digital
  • ICO Sandbox
  • Fairness
  • Explainability
  • Bias Auditing
PR
Product & Design
Hover to see skills
PR Product & Design
  • Product Ownership
  • Figma
  • Go-to-Market
  • Architecture Design
  • Cognitive Linguistics
Career

Professional Experience

2023 – Present
Founder & AI Engineer
RLYXA Limited, London, UK
  • Led end-to-end architecture of specialized AI-driven SaaS platforms and data pipelines, bridging advanced ML with strict regulatory compliance.
  • RLYXA SWRx: Built an AI-powered financial crime compliance platform generating a synthetic AML/CFT training dataset of 5M+ JSONL instruction-output pairs covering structuring, PEP screening, TBML, sanctions, and SAR narrative generation.
  • CLIHub (clihub.co.uk): Developed a full-stack cognitive linguistics SaaS with five AI-powered tools (CMT Workbench, Frame Grapher, CxG Debugger, Bias Auditor, Interpretability Dashboard). Led university partnership strategies targeting Lancaster, UCL, Birmingham, and Edinburgh.
  • CONNECT: Architected a secure multi-agency safeguarding AI system, currently in the ICO Regulatory Sandbox engagement phase.
  • RLYXA DPP: Developed a digital health prevention application with Imperial College Health Partners (ICHP).
2025 – Present
E-commerce Founder & Developer
Lekoz (lekoz.com), United States
  • Designed and launched an online pet retail business, managing end-to-end operations.
  • Built full-stack web development and an AI-powered digital marketing content strategy to drive customer acquisition.
2018 – 2024
Software Developer
Kayan HR, Amman, Jordan
  • Developed and maintained robust software solutions for HR management, ensuring high availability and performance.
  • Collaborated with cross-functional teams to implement new features, optimize database queries, and improve system architecture.
2006 – 2017
Software Developer
Housing Bank, Amman, Jordan
  • Engineered and deployed mission-critical financial software applications, ensuring strict adherence to banking security standards and data integrity.
  • Led the migration of legacy systems to modern architectures, significantly reducing processing latency and improving user experience.
Portfolio

Selected AI Projects

AI / ML

AML Fine-Tuning Dataset Pipeline

Engineered a large-scale synthetic training dataset across FATF categories (structuring, PEPs, TBML, crypto AML, sanctions, SAR narratives). Designed for supervised fine-tuning of open-weights LLMs for financial crime compliance.

Python JSONL LLM Prompting FATF/AML
RAG

RAG-Powered Internal Knowledge Assistant

Built a retrieval-augmented generation system utilizing LangChain, FAISS vector store, and GPT-4/Claude for internal policy Q&A. Implemented responsible AI guardrails including source citation and content filtering.

LangChain FAISS OpenAI API Claude API Docker
SaaS

CLIHub — Cognitive Linguistics AI Platform

Full-stack SaaS with five AI-powered cognitive linguistics tools and a self-contained 11-module, 32-lesson interactive HTML course. Produced formal reports for UK university outreach (Lancaster, UCL, Birmingham, Edinburgh).

React Node.js OpenAI/Claude Python
Visit CLIHub
Compliance

RLYXA SWRx — Financial Crime Compliance

AI-powered platform generating 5M+ synthetic AML/CFT training pairs. Architected pipeline for fine-tuning open-weights LLMs covering structuring, PEP screening, TBML, sanctions, and SAR narrative generation.

LLM Fine-Tuning LoRA/QLoRA AML/CFT Synthetic Data
Health

RLYXA DPP — Digital Health Prevention

Digital health prevention application currently in development with Imperial College Health Partners (ICHP). Focused on proactive health management and prevention through AI-driven insights.

Health AI NHS Digital Product Design
Safeguarding

CONNECT — Multi-Agency Safeguarding AI

Secure multi-agency safeguarding AI system currently in the ICO Regulatory Sandbox engagement phase. Sovereign open-weights AI strategy for post-approval deployment.

Regulatory AI ICO Sandbox Sovereign AI
FinTech / ML

Credit Default Risk Prediction

Built an XGBoost credit default prediction model with Bayesian hyperparameter tuning (Optuna), achieving 0.772 AUC-ROC. Integrated TreeSHAP for full explainability per FCRA/GDPR fair-lending requirements. Includes demographic parity checks and probability calibration.

XGBoost Optuna TreeSHAP Python Scikit-learn
Time-Series / Deep Learning

Hybrid Prophet-LSTM Demand Forecasting

Walmart M5 competition-level demand forecasting using a hybrid architecture: Prophet decomposes trend/seasonality, then a PyTorch Attention LSTM models the residuals. Evaluated with WRMSSE, MASE, RMSE, and MAE. Rich feature set includes SNAP calendars, holiday markers, and price elasticity.

PyTorch Prophet LSTM Attention Pandas
Education

Academic Background

🎓
BSc in Computer Science
University of Jordan
Class of 2005

Let's talk.

Open to AI engineering, full-stack, and product consulting opportunities.

✉️ mosheikh83@outlook.com 📱 07477 500574 🌐 rlyxa.co.uk 🌐 clihub.co.uk