Portfolio surface

AI systems, small products, and the craft around them.

A living portfolio of production GenAI, multi-agent orchestration, AI enablement, governed systems, and side projects. The CV is here as source material, but the page is meant to be explored like a workshop.

Download CV PDFSend a note
Mohamad Amin Al Hajj
Utrecht-based AI builder: production systems, governance, enablement, and experiments that leave useful traces.
Source-grounded portfolio assistant
MA

MAAH Chatbot

Ask about shipped systems and choices

live AI chatbot
Open the portfolio assistant for a larger conversation.
Start anywhereTry a systems question, a project question, or one of the suggested prompts below.
MAAH Bot

I am Mohamad's source-grounded portfolio assistant. Ask about production AI systems, product judgement, architecture, governance, patents, publications, or the path behind the work.

Architecture

Reference GenAI Platforms

Reusable blueprints for RAG and multi-agent assistant systems: ingestion, chunking, embeddings, vector retrieval, guarded generation, tool routing, evaluation, and cost/latency controls.

Shipped system

Product Assistant

Customer-facing multi-agent assistant for L1/L2 technical questions using open SKF data, SOTA RAG, and an orchestration layer developed with Harness engineering tools.

Team capability

AI Development Enablement

Internal enablement around Codex, Claude Code, GitHub Copilot, MCP servers, reusable skills, and state-of-the-art implementation patterns.

Infrastructure

Sovereign AI Exploration

Collaboration with Sferical AI on local GPU clusters in Sweden, NVIDIA 300-series GPU racks, open models, and NVIDIA Run:ai for sensitive enterprise workloads.

Good first question

Ask what Mohamad has shipped, how the systems are put together, or which proof points connect across projects.

Best use

Use the assistant as a source-grounded index for projects, architecture choices, tradeoffs, and concrete evidence.

100k+external AI assistant queries per month
28 secaverage response time after launch
35%better optimized alternatives accuracy
5SMEs led across AI engineering and data science

Production AI

Customer-facing GenAI, RAG, multi-agent orchestration, evaluation, monitoring, and delivery from requirements to production.

AI Enablement

Practical training, ambassadors, coding-agent workflows, reusable skills, and adoption patterns that help teams actually use modern tools.

Governed Systems

AI Act readiness, Credo.ai workflows, access controls, monitoring, and sovereign AI exploration for confidential workloads.