About
At Hawaz, a Riyadh consultancy that serves Saudi government entities, I automate the consulting work itself. Jobs that used to mean days of reading Arabic documents and filling spreadsheets now run end to end: my systems ingest hundreds of pages, score them against official frameworks, and hand the team a finished draft with the evidence behind every finding. Consultants spend their time on judgement and the client instead of data entry.
What that looks like in practice: an assessor that scores a government platform across all 20 axes of the DGA Digital Experience Maturity Index and checks itself against the official results; Rasid, which tracks 8,700+ Saudi government tenders, extracts their evaluation criteria, and ranks every opportunity so the bid team knows where to compete; and a review agent that reads a full contract or proposal in one pass and catches contradictions, date errors, and conflicting figures before the client does.
These systems are built to be trusted in government work. Code computes every number, the model does the judging, and any finding it can't quote from the source is dropped. They run on the firm's own GPUs with Saudi PII anonymized, so confidential client documents never go to a third-party cloud.
Before Riyadh I lived in the Netherlands, where I completed an M.Sc. in Applied Data Science at Utrecht University, wrote my thesis on satellite imagery at Readar, analyzed logistics data at Coty, and freelanced on data pipelines and scrapers. I'm British and Irish, and a native speaker of English and Arabic.
Experience
-
2026 — Present
AI Engineer & Consultant · Hawaz
Build production LLM tools for government-sector consulting: an automated DGA maturity assessor, a local document review agent with Saudi PII anonymization, and the Rasid tender-intelligence platform. Also shipped the bilingual LEAP 2026 booth site.
-
2022 — 2026
Freelance Programmer & Analyst · Mek Tech Services
Data pipelines and lifecycle evaluation for a vertical-farming ML product, scrapers for EU trade registries, and a RAG-based tool that tracks sentiment in think-tank publications over time.
-
2025
M.Sc. Thesis Researcher · Readar
Developed a method to normalize colour differences across multi-sensor satellite and aerial imagery, preparing the data for deep-learning super-resolution.
-
2023 — 2024
Logistics Finance Consultant · Coty
Analyzed SAP logistics data and built cross-departmental Power BI dashboards for KPI monitoring and launch planning.
-
2021
Flutter Developer · Fair Online BV
Built UI components and Firebase-backed booking workflows.
-
Education
M.Sc. Applied Data Science · Utrecht University, 2025
B.Sc. Data Science and AI · Maastricht University, 2024
Projects
-
Hawaz
Local Document Review Agent
Reads whole Arabic and English documents on an NVIDIA DGX Spark, anonymizes Saudi PII with reversible pseudonyms, runs deterministic cross-checks on dates and figures, and drops any finding whose quote it can't find in the source. I benchmarked 12 Arabic and English NER models to choose the anonymizer.
-
Hawaz
Digital Experience Maturity Assessor
Automates assessment against the DGA Digital Experience Maturity Index. Code computes every metric; an LLM judge assigns calibrated maturity bands, scored against official DGA results.
-
Hawaz
Rasid Tender Intelligence
Tracks 8,700+ tenders from Etimad and international portals, extracts evaluation criteria, and scores each opportunity with an explained rationale. Shipped through 90+ reviewed pull requests.
-
Freelance
Syrian Dialogue Center Analysis Platform
Scraped Arabic and English documents and tracked how conflict narratives shifted over time, using Qwen models for analysis and ChromaDB for retrieval.
-
Freelance
Secure University RAG Chatbot
A RAG chatbot over university data with role-based access for the public, students, and faculty, plus audit logging.
-
Freelance
BIGss Service Broker Platform
A Flask REST API that connects foreign companies with Saudi service providers, with PostgreSQL on Railway and a React frontend on Vercel.