Software engineer specialised in Agentic AI and autonomous workflows, holding a National Engineering Degree in Computer Science (ENSI, EUR-ACE and ASIIN accredited, with Honours, 2024) and a Research Master's in Advanced…
Software engineer specialised in Agentic AI and autonomous workflows, holding a National Engineering Degree in Computer Science (ENSI, EUR-ACE and ASIIN accredited, with Honours, 2024) and a Research Master's in Advanced Intelligent Systems (with Honours). Currently AI Engineer at Proxym, building AI agents for the banking sector: 4 autonomous agents deployed to production, 30+ custom tools and multilingual RAG pipelines (English/Arabic). Previously Full-Stack Developer at Catharsis, where he led a team of 3 engineers to deliver a cross-platform CRM from prototype to pilot. Certified by NVIDIA in AI for Predictive Maintenance.
<p>Developed and deployed 4 autonomous AI agents: 2 banking workflow automations and 2 customer-facing chatbots that execute complete banking tasks.<br>
Engineered 30+ custom tools and robust information workflows, with interruption management and middleware layers that ensure accuracy and safety for high-risk transactions.<br>
Architected multilingual RAG pipelines (English/Arabic) and ran retrieval rate benchmarks to optimise document extraction and synthesis.<br>
Drove the full development lifecycle, from prototype to production deployment, within an agile squad of 4 developers.<br>
Stack: LangChain, LangGraph, DeepAgents, RAG, FastAPI, Python</p>
<p>Led the end-to-end development of a new CRM product for medical sales representatives, taking it from prototype to a live pilot with a key client and 5+ representatives.<br>
Architected a secure, scalable cross-platform solution with React Native (TypeScript), Electron, Flask and PostgreSQL.<br>
Led a team of 3 engineers through the Agile/Scrum lifecycle, running sprint planning and code reviews and turning business needs into technical specifications.<br>
Set up secure backend hosting via Cloudflare Tunnels to ensure reliable remote access.</p>
<p>Implemented an unsupervised deep learning solution for anomaly detection in banking event logs, adapting a model from the scientific literature.<br>
Identified a critical failure in detecting timestamp-shift anomalies and traced its root cause to information loss in the convolutional encoder-decoder during node embedding aggregation.<br>
Ran a comparative study of model parameters that quantified an inverse relationship between temporal and structural anomaly detection.<br>
Built the full-stack implementation with Flask, React and PostgreSQL (Dockerised).</p>
نظام التوظيف التشغيلي المعتمد على الذكاء الاصطناعي أولاً، الذي يوحّد ويؤتمت دورة التوظيف بالكامل للمؤسسات الحديثة حول العالم.
د.جوب منطقة حرة ذ.م.م. 2026 © جميع الحقوق محفوظة
