The voyage — Hakeem Hannoon

Hi, I’m Hakeem, and I love Physics and AI. I want to be useful to other humans, understand more about the universe, and help answer some of the big questions.

Write to me See the work

Scroll

Work and Research

Jan – Jun 2026 Remote

Research Fellow

Algoverse AI Research Program
Show detailsHide details: Research Fellow, Algoverse AI Research Program
- Developed new ways to organize stored user memories, reducing memory misuse by language models by 8.8%.
- Tested the new memory methods on 10 language models using Azure and Google Cloud.
- First author of the May 2026 preprint “Mitigating Over Personalization in Language Models via Structured Memory.”
- Python
- AI research
- Language models
- Azure
- Google Cloud
- Model evaluation
- Research writing
Read the paper
May – Oct 2025 Saskatoon, SK

Big Data Analyst

Advanced Physics Research Lab, University of Saskatchewan
Show detailsHide details: Big Data Analyst, Advanced Physics Research Lab, University of Saskatchewan
- Built a Python quantum chemistry model to rank 25 chemical molecules by toxicity.
- Improved the previous model’s ranking by 31%.
- Built Python and C tools to move large research datasets from supercomputers to the cloud for analysis and processing.
- Python
- C
- Data analysis
- Cloud computing
- Quantum chemistry
- Scientific computing
May – Oct 2024 Saskatoon, SK

Research Assistant

Department of Mathematics & Statistics, University of Saskatchewan
Show detailsHide details: Research Assistant, Department of Mathematics & Statistics, University of Saskatchewan
- Developed algorithms in C, C++, and Python pipelines on supercomputers to simulate complex interacting-particle systems and capture their collision behavior in up to 100 dimensions.
- Built models that identified critical behavior with 95 percent accuracy in higher dimensions, matching theoretical predictions.
- Mathematical biology
- Diffusion systems
- HPC
- Simulation
- Algorithm design
- Scientific computing
- Python
- Applied mathematics
Jun – Aug 2024 Saskatoon, SK

Research Assistant

quanTA, University of Saskatchewan
Show detailsHide details: Research Assistant, quanTA, University of Saskatchewan
- Worked on foundational quantum information theory and near-term quantum computing applications.
- Implemented Qiskit code for quantum algorithms and IBM Qiskit Summer School projects.
- Explored quantum processing applications in materials discovery, health sciences, and energy.
- Quantum computing
- Qiskit
- Quantum information
- Python
- IBM Qiskit
- Scientific computing

Publications

2026 Paper

Mitigating Over-Personalization in LLMs via Structured Memory

First author. Accepted to AACL-IJCNLP 2026 (Hengqin, China, November 2026) and to multiple ICML workshops.
Show detailsHide details: Mitigating Over-Personalization in LLMs via Structured Memory
This paper studies safety failures in memory-augmented conversational LLMs, especially cross-domain leakage and memory-induced sycophancy. It proposes inference-time structured memory representations that partition memories by domain instead of injecting all stored user memories as a flat list. Across seven models on PersistBench, the strongest dynamic partitioning method reduces cross-domain leakage by 8.8 percent on average relative to the baseline while preserving useful personalization.
- LLM safety research
- Python & large-scale evaluation
- Research writing
Read the paper : Mitigating Over-Personalization in LLMs via Structured Memory Download PDF of Mitigating Over-Personalization in LLMs via Structured Memory arXiv:2608.08300 (opens in a new tab)
2026 Thesis

Honours Thesis

Author. Undergraduate Honours Thesis.
Show detailsHide details: Honours Thesis
Research on symmetry methods for solving non-linear partial differential equations — using the structure of a system's symmetries to construct exact and reduced solutions where direct integration fails.
- Symmetry methods for PDEs
- Mathematical modeling
- Scientific computing
Read the thesis : Honours Thesis Download PDF of Honours Thesis
2025 Paper

Sustainable Tourism in Juneau, Alaska: A Stochastic Multi-Agent Model

Co-author, with Cole Dutchyn (opens in a new tab) and Jason Main (opens in a new tab). 2025 Mathematical Contest in Modeling (MCM), Problem B, University of Saskatchewan.
Show detailsHide details: Sustainable Tourism in Juneau, Alaska: A Stochastic Multi-Agent Model
A model of sustainable cruise tourism in Juneau, Alaska, written over four days for the 2025 Mathematical Contest in Modeling. Taking its cue from RollerCoaster Tycoon, it follows every tourist's happiness through the day as they choose among the glaciers, whale watching, the tram and the city. Crowding and a tourism tax raise or lower that happiness, and happier visitors bring more ships in the days that follow. Raising the tax from 5 percent to 22.5 percent keeps the city's carbon footprint stable, while 15 to 20 percent gives the lowest carbon footprint per dollar of tourism revenue.
- Stochastic multi-agent modeling
- Simulation
- Sustainability policy analysis
Read the paper : Sustainable Tourism in Juneau, Alaska: A Stochastic Multi-Agent Model Download PDF of Sustainable Tourism in Juneau, Alaska: A Stochastic Multi-Agent Model

Projects

Spotter

Got tired of fighting clunky forms to log my workouts, so I built an app that logs everything for me when I just talk to it.

Visit spotter-labs.com (opens in a new tab)
Aether Field

I really liked Windows XP media player as a kid, so wanted to make a new one.

Visit hakeem-hannoon.github.io/aether-field (opens in a new tab)
RAGuccino Nutrition Scout

Built a RAG-powered nutrition scout so retrieval can chase new Starbucks drinks before static food tables wake up.

View the code on GitHub (opens in a new tab)