Engineering intelligent cloud systems.
Hi, I'm Qossay Abuamriya. A Cloud & AI Software Engineer with 5 years of experience, based in Dubai. I specialize in Agentic RAG, LLM fine-tuning, NLP, OCR, and highly available Kubernetes—leveraging Azure, OpenAI, IBM, and Cohere to maximize AI impact.
I build production systems that combine AI, cloud-native architecture, and strong engineering practices—so teams can ship faster and scale with confidence. From requirement analysis to deployment and monitoring, I focus on control, security, and measurable outcomes.
About me
I'm a versatile AI and cloud expert who brings a blend of technical depth and business understanding to AI teams. My experience spans from being the AI provider at CoolRiots, where I built and optimized AI-driven platforms, to leading business use cases at Takaful Emarat, where I implement AI solutions tailored to real-world challenges. This dual perspective helps me understand requirements and translate them into implementations that deliver tangible value.
Over the past 5 years I've focused on innovation through AI, cloud, and automation. At CoolRiots (IBM Partner), I built the company's DevOps infrastructure from scratch for scalable growth and seamless deployments. As an AI Product Engineer I developed the BeX platform—empowering businesses to create AI agents, chatbots, and workflows for industry-specific challenges. Now at Takaful Emarat in Dubai, I lead digital transformation by implementing AI-powered solutions that streamline operations and enhance customer experiences. My aim is to bridge cutting-edge technology and real-world business needs.
I hold a degree in Software Engineering from the University Malaysia of Computer Science and Engineering (UNIMY), in Cyberjaya, Malaysia's tech hub. My career has taken me through Kuala Lumpur, Singapore, and now Dubai.
Featured Work & Impact
Key projects from Takaful Emarat & CoolRiots (IBM Partner)—delivering measurable impact on efficiency, accuracy, and reliability.
Agentic AI Customer Support
Developed an AI-powered customer support agent that generates tailored, well-structured responses from user data and is equipped with tools to interact with core system functionalities. Built on the Airia Platform with LLMs from OpenAI and Cohere; implemented guardrails, content filtering, and Responsible AI practices for safe, reliable interactions—cutting turnaround time by 70%.
BeX & Agentic Business Platform
As AI Product Engineer, developed the BeX platform—a cloud-native, microservice-based Agentic AI product in Python that lets businesses create AI agents, chatbots, and workflows for hospitality, law, and education. Fine-tuned LLMs on IBM WatsonX Studio and Azure OpenAI; used RAG with vector DB and LangChain. Built company DevOps from scratch; integrated Groq APIs for faster inference.
Hybrid Search & RAG
Designed and implemented AI-powered search workflows with data refinement, hybrid search (keyword + semantic), and retrieval from SQL, unstructured text, and vector stores using LangChain, Azure AI Search, and AI Foundry. Integrated Cohere Command R+ and Embed v4 into Agentic RAG pipelines—improving older RAG accuracy by 40% and cutting data retrieval time by 30%.
Identity Document OCR
Led requirements and built a cost-efficient internal tool for identity document processing using Azure OCR, Document Intelligence, and Azure OpenAI LLMs. Handles mixed document types, formats, and resolutions; optimized data refinement and model selection to cut processing time by 50% and feed structured data into downstream workflows.
Claims Management App
Full-stack app with Next.js, FastAPI, and SQL Server. Built with best-practice stored procedures, role-based profiles and authorization, Azure Authentication, and SSO for secure, scalable claims handling.
Deployment & Time-to-Market
Use Azure App Service and Container Apps to deploy and configure applications; set up CI/CD pipelines and ingress controllers so new features reach production faster—reducing time-to-market by 25%.
Cloud Infrastructure & K8s
Implemented Kubernetes (AKS, IKS, OpenShift) for application orchestration with NGINX load balancing and high availability per IBM and Azure standards. Delivered 20+ CI/CD pipelines (Azure DevOps, GitHub Actions), Terraform for Azure and K8s, and Grafana with Loki and Prometheus for monitoring—achieving 99.95% uptime and ~30% faster deployments and setup.
Technical Arsenal
Tools and practices I use daily to design, build, and operate reliable systems.
Generative AI & ML
From prompt design to RAG and fine-tuning for production.
Cloud & DevOps
Infrastructure as code, orchestration, and observability.
Software Engineering
APIs, data layer, and iterative delivery.
Languages: Arabic (native), English (professional).
Certifications & Education
Formal credentials and continuous learning in cloud, AI, and DevOps.
Microsoft Azure
- Azure AI Engineer Associate Jan 2025
- Azure Developer Associate May 2025
- Azure Administrator In progress
- Azure DevOps Expert In progress
- Azure AI Fundamentals May 2023
- Azure Security, Compliance & Identity Mar 2023
- Azure Data Fundamentals Feb 2023
DevOps & IBM
- Certified Kubernetes Administrator (CKA) Feb 2024
- IBM Watson Assistant Sales Foundation May 2023
- IBM Watson Assistant Technical Sales May 2023
Education
Bachelor’s in Software Engineering (Hons)
University Malaysia of Computer Science & Engineering (UNIMY), Kuala Lumpur • 3.73 CGPA • Full scholarship • Mar 2024