Data Scientist with experience producing tools for open-source intelligence analysis centered on natural language processing (NLP), visualization, and mapping tasks. Develops and fine-tunes supervised and unsupervised ma…
Data Scientist with experience producing tools for open-source intelligence analysis centered on natural language processing (NLP), visualization, and mapping tasks. Develops and fine-tunes supervised and unsupervised machine learning models for topic modeling and sentiment analysis, and designs experiments and evaluation frameworks for assessing upstream techniques and model performance. Previously conducted data analysis and qualitative research to produce dark web intelligence for client financial institutions and produced qualitative cybersecurity threat reports for non-technical stakeholders.
I am a data scientist currently working at an open-source intelligence center in Abu Dhabi. From my experience in threat intelligence and cyber risk management, I provide a unique data-driven perspective that leverages my cyber expertise along with research, writing, and data science skills. I am currently interested in opportunities to produce innovative analysis and solutions in this space.
Produce tools to enable open-source intelligence analysis centered on natural language processing (NLP), visualization, and mapping tasks
Develop and fine-tune supervised and unsupervised machine learning models for topic modeling and sentiment analysis
Design experiments and evaluation frameworks for assessing upstream techniques and model performance
Review literature on upstream NLP methods (language transformers, tokenizers, pre-processing steps etc.)
Used Tensorflow, Keras, scikit-learn, and Natural Language Toolkit to train a neural network for English sentiment analysis, improving the accuracy of legacy in-house SVM models by 5%
Used Tensorflow, Keras, and Transformers to create a calibrated open-source BERT model for sentiment analysis
Used Top2Vec to fine-tune BERT-based models for domain-specific topic modeling use cases in Arabic and English
Used Spacy to create tokenizers leveraging named entity and noun phrase recognition
Identified optimal configurations for specific domain and language combinations
Document components of the model development process (literature reviews, background information/justification for upstream techniques, experiment results for performance on downstream tasks)
Research and evaluate open-source geoparsing and geolocation tools (Geoparsepy, Osm-Python, Photon, Nominatim)
Research and implement statistical methods with SciPy for evaluating performance
Used ElasticSearch, Airflow, PostgreSQL, and Tableau to create an ETL pipeline for a dashboard reporting aspect-based sentiment by topic and location
Work with analyst teams to identify new lines of research and analysis
Advise teams on leveraging available data sources and in-house models to fulfill intelligence requirements
Automate analyst workloads by creating tools for translation workflows (Django REST, Docker, reverso), visualization (plotly, seaborn), mapping (geopandas, keplergl), scraping (beautiful soup), and parsing tasks
Produce tools for open-source intelligence analysis using NLP, visualization, and mapping tasks
Develop and fine-tune machine learning models for topic modeling and sentiment analysis
Review literature on NLP methods
Train a neural network for English sentiment analysis
Create a calibrated open-source BERT model for sentiment analysis
Fine-tune BERT-based models for domain-specific topic modeling
Identify optimal configurations for domain and language combinations
Research and evaluate geoparsing and geolocation tools
Create an ETL pipeline for dashboard reporting aspect-based sentiment by topic and location
Work with analyst teams to identify new research lines and advise on data sources and models
Automate analyst workloads by creating tools for various tasks
Conducted data analysis and qualitative research to produce dark web intelligence for client financial institutions, with a focus on financial data breach activity
Used scikit-learn to develop a Random Forest model to detect abnormal transaction activity and cards at risk of fraud, replacing time-consuming excel-based analysis and achieving a recall score of 98%
Produced a weekly client-facing report analyzing the implications of recent cybersecurity and fraud trends affecting the finance sector
Used Python to analyze dark web shop activity and transaction data to identify cybercriminal patterns and new breach activity
Visualized criminal marketplace trends (matplotlib)
Automated data collection processes
Produced ad hoc long-form research reports analyzing broad trends affecting the cybercrime landscape
Provided bespoke reporting and analysis following client-specific incidents
Prepared materials for law enforcement upon request
Presented research findings at external conferences
Used security tools such as VirusTotal, SecurityTrails, and URLScan to identify malware campaigns and e-commerce breaches
Conducted client touchpoint meetings
Prepared materials briefing clients on dark web terminology and the fraud landscape
Produced qualitative reports providing an overview of cybersecurity threats, threat actors, vulnerabilities, and risk management strategies for non-technical board members and C-suite executives
Researched cyber risk management principles and relevant studies on the business impact of cybersecurity incidents and regulatory measures
Produced in-depth case studies concerning the causes and costs of major cyber incidents
Conduct data analysis and qualitative research for dark web intelligence
Develop a Random Forest model to detect abnormal transaction activity
Produce client-facing reports on cybersecurity and fraud trends
Analyze dark web shop activity and transaction data
Produce ad hoc research reports on cybercrime trends
Provide bespoke reporting and analysis for client-specific incidents
Present research findings at external conferences
Identify malware campaigns and e-commerce breaches
Conduct client touchpoint meetings
Produce qualitative reports on cybersecurity threats and risk management
Research cyber risk management principles and business impact of cybersecurity incidents
Produce in-depth case studies on cyber incidents
نظام التوظيف التشغيلي المعتمد على الذكاء الاصطناعي أولاً، الذي يوحّد ويؤتمت دورة التوظيف بالكامل للمؤسسات الحديثة حول العالم.
د.جوب منطقة حرة ذ.م.م. 2026 © جميع الحقوق محفوظة
