Abdullah Shaar
Verified Expert in Engineering
AI Engineer and Developer
Doha, Qatar
Toptal member since May 14, 2026
Abdullah is an AI engineer with a dual bachelor's degree in computer science and biological sciences from CMU. At Qatar Computing Research Institute, he fine-tunes large language models (LLMs), including Fanar, and builds genomic risk models for precision medicine, shipping end-to-end pipelines and production tools across both domains. An Oxford ML Summer School graduate, Abdullah is open to ML, LLM, bioinformatics, and data engineering projects.
Portfolio
Experience
- Biostatistics - 6 years
- Large Language Models (LLMs) - 5 years
- PyTorch - 5 years
- Deep Learning - 5 years
- Artificial Intelligence (AI) - 5 years
- LangGraph - 4 years
- Google Cloud Platform (GCP) - 3 years
- AI Agents - 3 years
Preferred Environment
Python 3, PyTorch, LangGraph, AI Agents
The most amazing...
...thing I've built is an AI scheduler deployed at HMC, Qatar's largest hospital that automates residency rotation planning for 80+ doctors.
Work Experience
AI Research Engineer
Qatar Computing Research Institute
- Engineered a self-evolving GRPO training and evaluation pipeline for Fanar, Qatar's first Arabic multimodal LLM, enabling autonomous cultural alignment of AI-generated images through dynamically updated evaluation rules.
- Built a transformer-based meta-learning system integrating multiple predictive models for biological data analysis, validated on 2 international cohorts, including the UK Biobank, and published in the Journal of the American Heart Association.
- Built an autonomous genomics analysis agent using LangChain and LangGraph powered by Mixtral, replacing manual bioinformatics workflows with fully orchestrated multi-step pipelines.
AI Engineer
Avey
- Fine-tuned a BERT-based NLP model to extract clinical symptoms from unstructured medical text, achieving 98% accuracy and enabling automated patient intake processing at scale. It is currently deployed by AveyAI.
- Developed and operationalized ML pipelines for health insurance fraud detection, improving data integrity and reducing financial risk exposure across thousands of insurance claims.
- Built and deployed interactive React dashboards to visualize ML predictions and fraud detection outputs, enabling non-technical stakeholders to monitor model performance in real time.
Experience
Medical Residency Rotation Scheduler
The system uses Google OR-Tools CP-SAT, a complete constraint programming solver, to globally optimize annual rotation assignments for 60–80 residents across 22 clinical rotations and 13 scheduling blocks. The model encodes 11 categories of hard constraints, including PGY-level graduation requirements, minimum staffing levels, leave enforcement, sequential rotation rules, and batch integrity, alongside 8 weighted soft constraints that maximize schedule quality via an objective function.
It is architected as a modular Python package with 3 independent interfaces: a Streamlit web app for interactive use, a CLI with full argparse support, and an ipywidgets-powered Jupyter notebook for exploratory analysis. Output is a multi-sheet Excel workbook with per-PGY schedules, staffing summaries, and a full constraint satisfaction log.
Education
Dual Bachelor's Degree in Computer Science and Biological Sciences
Carnegie Mellon University - Pittsburgh, PA, USA
Skills
Libraries/APIs
PyTorch
Languages
Python 3, Python, R, C, Java
Frameworks
LangGraph
Platforms
Google Cloud Platform (GCP)
Other
Machine Learning, Deep Learning, Large Language Models (LLMs), Biostatistics, AI Agents, Artificial Intelligence (AI), Optimization, Retrieval-augmented Generation (RAG), Architecture
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring