Author Image

Hi, I am Mohammed

Mohammed Nagdi

Software Engineer at Zain

I’m an AI and Machine Learning enthusiast with an M.Sc. in Artificial Intelligence from the University of Edinburgh and awarded the Google DeepMind Scholarship. I specialise in Reinforcement Learning (RL) and Natural Language Processing (NLP) and have a strong background in time-series forecasting, recommender systems, and system identification.

Driven by a passion for AI since my undergraduate studies, I co-organized Indabax conferences in Sudan and pursued my master’s to apply technical advancements in AI. With experience as a Software Engineer at Zain, I engineered data integration solutions and developed APIs processing over 1M+ daily transactions. My research includes time-series foundational models, reinforcement learning, and system identification using Koopman theory and GPT architecture.

I aim to contribute to impactful AI projects, leveraging my expertise in ML, NLP, and software engineering to drive innovation and ethical AI practices.

Experiences

1
Zain

March 2023 - Present

Remote

Middleware Development Team Leader

Feb 2026 - Present

Responsibilities:
  • Designed and deployed a production recommendation system integrating collaborative filtering, subscription history, and pricing guardrails — driving 8% revenue uplift across 2–3M daily customers via Meta’s Upselling APIs.
  • Led middleware engineering team across API infrastructure processing millions of daily transactions; established architecture standards, CI/CD workflows, bank payment gateway, and observability practices (Datadog, Dynatrace POCs).
Sr. Software Engineer

March 2023 - Jan 2026

Responsibilities:
  • Architected and maintained large-scale API infrastructure using WSO2 API Manager; engineered data integration solutions across BSS/OSS telecom ecosystem with CI/CD pipelines (GitLab) and container registries
  • Collaborated with AWS, GitLab, WSO2, and Datadog teams to resolve integration challenges and optimise performance across distributed environments
University of Edinburgh

Jan 2025 - Dec 2025

London, United Kingdom

Research Assistant

Jan 2025 - Dec 2025

Responsibilities:
  • Lead research collaboration with the Kutz Research Group (University of Washington) on dynamical systems approaches for synthetic biology, addressing challenges beyond traditional models for time series forecasting.
  • Applied advanced methodologies, including Koopman operator, Dynamic Mode Decomposition (DMD), SHRED, SINDy architecture, and Kolmogorov-Arnold Networks (KAN), to model physical and biological systems like the repressilator and Duffing oscillator.
2
3
QAST

Oct 2021 - Mar 2023

Khartoum, Sudan

SRE - Enterprise Applications

Oct 2021 - Mar 2023

Responsibilities:
  • Implemented automated monitoring solutions using Centreon and custom Nagios scripts, enhancing system availability by 30% through proactive issue detection and resolution
  • Designed analytical dashboards using Apache Superset, delivering actionable KPI insights that improved IT operations productivity by 10% and reduced report generation time by 40%
Business Associates

Feb 2021 - Aug 2021

Khartoum, Sudan

Business Associates helps organizations to improve their performance, primarily through the analysis of existing organizational problems and development of plans for improvement

Project Manager

Feb 2021 - Aug 2021

Responsibilities:
  • Gathering product requirements, prioritising feature implementations, setting up development plan and improving user experience.
4

Education

University of Edinburgh
2023-2024
M.Sc. in Artificial Intelligence
Grade: Merit out of
Publications:
Taken Courses:
  • Reinforcement Learning
  • Case Studies of AI Ethics
  • Applied Machine Learning
  • Machine Learnign Practical
  • Accelerated Natural Language Processing
  • Natural Language Understanding, Generation, and Machine Translation
  • Informatics Project Proposal
  • Informatics Research Review
  • Applied Cloud Computing
Funding:
Google Deepmind Scholarship
University of Khartoum
2014-2020
B.Sc. in Electrical and Electronics Engineering
CGPA: 8.17 out of 10
Publications:
Extracurricular Activities:
  • IndabaX Sudan
  • Academic Committee for the (Electrical and Electronic Engineering Student Exhibition)

Projects & Publications

Building a GPT for System Identification
MSc Thesis • University of Edinburgh Aug 2024

GPT + Koopman + KAN model that beats Chronos, MOMENT, TimesFM on multivariate dynamical systems.

LLM Koopman KAN Time-Series
Details
Robust Identification of Synthetic GeneNetworks via Koopman Autoencoders with Latent Memory
Under review at IFAC • First author Nov 2025

AFT memory layer + Dynamical Re-encoding for robust synthetic gene network identification.

Koopman Autoencoders Synthetic Biology
Company SIC Classification & Summarisation Pipeline
Production NLP System Jan 2025

Fine-tuned DistilBERT (multi-label SIC) + T5/LLaMA summarisation on crawled UK company data.

NLP Fine-tuning LLaMA RAG
Apache Superset Analytical Dashboards
Enterprise Engineer • QAST 2021 – 2023

Built KPI dashboards that cut report time 40% and boosted IT productivity 10%.

Apache Superset Analytics BI Data Visualization
Ericsson → Huawei Middleware Portfolio Rebuild
Platform Engineer • Zain 2023 – 2024

Led full WSO2 middleware migration and rebuild under tight deadlines.

WSO2 Migration Leadership Enterprise
Multimodal Transformer Fusion for Alzheimer’s Detection
Research Project • University of Edinburgh Apr 2024

CNN+UNet encoders → transformer fusion of PET/MRI/CT. Outperformed uni-modal baselines.

Medical Imaging Multimodal Transformers
Deep RL for Robotic Hand Manipulation
IEEE ICCCEEE 2021 Paper Jan 2021

Benchmarked DDPG, DDPG+HER, PPO on Fetch & Shadow Hand tasks.

Reinforcement Learning Robotics PPO
Details

Skills

PyTorch
PyTorch

My go-to framework – I build and train everything in PyTorch: custom Time-series models, RL agents, multimodal transformers, and fine-tuned LLMs.

Large Language Models & Agents
Large Language Models & Agents

Beginner practitioner exploring LLMs with Hugging Face Transformers and LangChain. Built simple RAG pipelines, tool-calling agents, and tracked experiments using Weights & Biases.

Data Science & Classical ML Toolkit
Data Science & Classical ML Toolkit

Daily driver: Pandas, NumPy, Scikit-learn, Matplotlib + Optuna tuning and Gymnasium RL environments.

Middleware & API Integration
Middleware & API Integration

I architect and maintain enterprise-grade API platforms with WSO2 API Manager, build custom Java connectors, and design RESTful, GraphQL & SOAP APIs with Swagger/OpenAPI.

Database & Data Platforms
Database & Data Platforms

I run and tune Oracle, PostgreSQL, MySQL, SAP HANA & Superset in production – backups, replication, security, and dashboards.

CI/CD & Platform Automation
CI/CD & Platform Automation

I build and run production CI/CD with GitLab CI, GitHub Actions, Docker, and self-hosted runners – powering middleware environment at Zain.

Linux
Linux

Proficient in using Linux as the main operating system, with expertise in writing bash/shell scripts and developing auto-installation scripts.

Git & Workflows
Git & Workflows

I set up and maintain GitLab Enterprise workflows, protected branches, code reviews, and runners in production.

Fast Learner
Fast Learner

Non-CS background → Google DeepMind Scholar, ML Research publication, and production ML systems while working full-time as a Software Engineer.

Teamwork & Collaboration
Teamwork & Collaboration

I led the full Ericsson → Huawei middleware portfolio rebuild at Zain – coordinated cross-functional teams and external vendors to deliver under tight deadlines.

Leadership
Leadership

I’ve led teams of 5–20+ people – from running IndabaX Sudan to driving enterprise projects as Senior Engineer at QAST.