Profile
AI Solutions Architect with 18+ years spanning enterprise, solution, cloud, and AI architecture, including hands-on production delivery of generative AI, RAG, and agentic systems for large-scale, regulated organizations. I'm currently embedded within the Saudi Public Investment Fund (PIF) as part of a dedicated AI solutions team, taking GenAI use cases from discovery and ideation through PoC and production on Palantir Foundry and AIP — personally building RAG pipelines with vector, hybrid, and graph-based retrieval, agentic workflows orchestrated with AIP Logic, chatbot applications, and recurring LLM evaluation and regression-testing suites, while leading and mentoring a data engineering team. This flagship engagement caps a multi-sector NEOM consulting portfolio spanning Food, Environment, Energy & Water, Hotels, and the University of NEOM, across digital strategy, data strategy, business capability mapping, procurement, and system design. I'm skilled at translating ambiguous business problems into scalable, well-governed AI architectures while partnering with executive stakeholders. I recently completed an M.Sc. in Artificial Intelligence (University of Texas at Austin, GPA 3.93) and am certified across the full Google Cloud AI/data/platform stack, plus AWS Solutions Architect Professional and Palantir.
Core Competencies
Generative AI & Agentic Systems
LLMs, Retrieval-Augmented Generation (vector, hybrid & graph-based retrieval), agentic/autonomous workflows, prompt & context engineering, Palantir AIP Logic orchestration, ASR pipelines.
AI Engineering & Evaluation
LLM evaluation datasets & regression testing (retrieval relevance, groundedness, latency, cost), embeddings & vector search, model explainability (Grad-CAM, SHAP), MLOps & model monitoring, cost & latency optimization.
System Design & Architecture
Enterprise & Cloud Architecture, Reference Architectures, HLD/LLD, Integration & Data-Flow Design, NFRs, Secure Landing Zones, Cost/Option Analyses.
Architecture Governance
Technical Evaluation & Compliance Reviews, RFP Development, Vendor & Solution Assessment, BDAT Domains (Business, Data, Application, Technology).
Platforms & Cloud
Palantir Foundry & AIP, Google Cloud (Vertex AI, BigQuery), AWS, Python/PyTorch/Hugging Face, Docker & Kubernetes.
Leadership & Delivery
Engineering Team Leadership & Mentoring, Digital & Data Strategy, Executive Stakeholder Engagement, Procurement Advisory, PoC-to-Production Delivery.
Experience
AI & Cloud Solutions Architect — TONOMUS (NEOM)
Riyadh, Saudi Arabia · Oct 2021 – PresentEmbedded solutions consultant serving multiple NEOM sector clients across the full engagement lifecycle — from digital & data strategy through business capability mapping, procurement advisory, system design, and technology consulting — culminating in hands-on production AI delivery at PIF.
- PIF flagship delivery: architected and hands-on built an end-to-end enterprise GenAI system on Palantir Foundry & AIP — RAG pipelines, ASR ingestion, agentic workflows, RBAC, lineage, and telemetry — designed to production-grade NFRs (security, latency, cost, observability) in a regulated public-sector environment.
- Hands-on engineering ownership: personally built and owned production Foundry components — data pipelines, ontology objects, chatbot applications, RAG pipelines, and data visualizations — with agentic orchestration implemented in AIP Logic.
- Advanced retrieval: designed and built enterprise RAG pipelines combining vector, hybrid, and graph-based retrieval — source curation, chunking, embeddings, and prompt orchestration — for grounded, source-attributed answers.
- LLM evaluation & regression testing: defined evaluation datasets and recurring regression tests covering retrieval relevance, groundedness, latency, and cost.
- Engineering leadership: led and mentored the data engineering team of Saudi fresh-graduate engineers through design reviews and delivery ownership.
- Authored NEOM enterprise AI and cloud reference architectures, landing zones, HLD/LLD, integration architectures, and NFR & cost/option analyses across sector programs — Food, Environment, Energy & Water, Hotels, and University of NEOM.
- Advised on procurement and technology selection, and facilitated AI discovery workshops and executive presentations for senior client stakeholders.
Solution Architect — Mobily
Riyadh, Saudi Arabia · Jul 2017 – Sep 2021- Partnered with business and IT stakeholders to define, review, and refine requirements, ensuring alignment with architectural standards; led architecture presentations and workshops to walk through proposed solutions and gather governance feedback.
- Documented solution architecture artifacts — business use cases, data flows, functional/ non-functional requirements — and produced system diagrams identifying interdependencies and risks, across IT, network, and security domains.
Senior Consultant — Ericsson
Sulaymaniyah, Iraq · Jul 2013 – Jun 2017- Provided hands-on architecture support on a large-scale Telecom IT Transformation program for Asiacell Telecom, leading a development team customizing the BSCS iX R3 billing platform.
- Advised management on development estimates, technical risk, and architecture/technology choices; built integration endpoints between BSCS iX and CRBT, online mediation, POS, and CRM platforms.
Consultant — Atos
Rabat, Morocco · Aug 2012 – Jun 2013- Designed and implemented software applications for new products and offers for Maroc Telecom, including a Spring Batch application for tariff-model changes and an end-to-end location-monitoring service.
Software Engineer — Xpress Integration & Ericsson
Cairo, Egypt · Jul 2008 – Jul 2012- Designed and developed the SPT (Smart Promotion Tool) and a SQL query-builder module for customer-eligibility targeting, plus re-rating batch processing; earlier built GSM network-simulation and mass-service-activation tooling at Ericsson.
Education
M.Sc. in Artificial Intelligence — University of Texas at Austin
Sep 2024 – May 2026 · GPA 3.93/4.00- Selected project — Medical Imaging Diagnostic Pipeline: trained a brain-cancer classification model with structured error analysis and Grad-CAM/SHAP explainability, reducing false positives by 13%.
- Selected project — CNN vs. Vision-Language Models for Explainable Medical Imaging: compared CNN explainability methods (Grad-CAM++, EigenCAM, AblationCAM) with LLaVA-Med few-shot prompting on brain-tumor MRI; the VLM correctly classified 9 of the 13 scans the CNN misclassified (69.2%, 4-shot).
Postgraduate Diploma in Software Engineering — ITI
Cairo, Egypt · Jul 2007 – Jun 2008B.Sc. in Mathematics and Computer Science — Menoufia University
Shebin Elkom, Egypt · Sep 2001 – May 2005Certifications
Google Cloud
- Professional Machine Learning Engineer
- Professional Data Engineer
- Professional Cloud Architect
- Generative AI Leader
- Associate Cloud Engineer
AWS
- Solutions Architect – Professional
- DevOps Engineer – Professional
- Solutions Architect – Associate
- Cloud Practitioner
Palantir
- Foundry Data Engineer Associate
- Foundry Application Developer Associate
- Foundry & AIP Builder
Languages
- Arabic — Native
- English — Fluent