Aiman
Al-Hazmi.
MSc Computer Science · AI & ML Engineer · Software Developer.
About
A little about me.
I'm a Computer Science student focusing on Data Science and Machine Learning, with experience in software development and applied ML. My recent work is in agentic AI: multi-agent systems, retrieval-augmented generation, and the question of how to make an LLM system's answers verifiable rather than merely fluent. I enjoy taking research ideas and turning them into software that actually runs.
Outside of academics and tech, I'm also actively involved in community volunteering.
Selected work
Featured projects
MAScan
Multi-agent AI system for strategic market analysis. Six specialist agents, one per PESTEL dimension, each with its own evidence source.
Stratum — Open-Source Contributor
An open-source ML system for efficiently executing large-scale agentic pipeline search.
Thabetha
Mutually-confirmed debt tracking for shops, freelancers, and friends. Built at SalamHack 2026.
From Classical ML to LLMs in NLP
A comparative analysis of NLP evolution, evaluating multilingual transfer in medical contexts and benchmarking LLMs against classical machine learning on structured data.
Building Mini LLM
An end-to-end Mini-LLM pipeline that evolves from statistical n-gram models to neural and transformer-based language models, with systematic benchmarking and evaluation.
Financial News Sentiment Analysis
Sentiment classifier for financial headlines using Naive Bayes, FFNs, and BERT-based RNNs. F1 0.773 multiclass / 0.892 binary.
Fairfluence
Open-source Python tool for advanced dataset profiling and fairness analysis. Integrates with OpenML and Hugging Face.
Writing
Latest from the blog
Aug 1, 2026
LangGraph: Graph-Based Orchestration of Agents with Durable State and Explicit Control Flow
What I learned while building an agentic retrieval pipeline with LangGraph: why an explicit control flow is more reliable than agent autonomy, and why the recursion limit is not a loop counter.
Apr 7, 2026
Can We Trust AI Agents? The 'Human-on-Top' Necessity
An experiment using Claude as the architect and local models for routine coding — and what went right and wrong.