Alfarzdag Mohamed

• Provided expert advice to clients, identifying and solving data science challenges to enhance business efficiency.
• Designed and implemented machine learning models and AI solutions, including deep learning and NLP applications.
• Facilitated workshops to translate business problems into technical requirements, producing prototypes and project plans.
• Collaborated in an Agile environment

• Interactive Dashboard Development: Building user-friendly dashboards using Dash and R Shiny for stakeholder data visualization and self-service analytics.
• Stakeholder Engagement: Working directly with customers to gather requirements and translate business needs into technical solutions.
• Data Pipeline Development: Acquiring new data sources and building robust ETL pipelines to create compreh

• AI Chatbot Development: Led development of IA (Impact Assessment) chatbot using retrieval-augmented generation (RAG) and Large Language Models, improving regulatory quality assessment from 63% 'weak' ratings. Built tool accessing 5,000+ IAs with rapid summarization capabilities.
• R&D Spatial Data Platform: Spearheaded takeover and redesign of R&D spatial data visualization tool from Nesta, crea

Performed data source comparisons using Python and built data pipelines to automate joining different table types. Delivered statistical advice and analysis while working with 30–40 stakeholders monthly, using SQL-based systems and unit testing.

• Performing data sources congruences including admin and surveys data using python.
• Building data pipeline to automate joining different type of tables using payspark.
• Deliver high quality statistical advice and analysis.
• Work proactively with 30-40 stakeholders monthly to identify their requirements.
• Skills: Python, R, Pyspark, Hive, Impala, SQL and Unit Testing.

• Worked on Fake News Detection Project (NLP) with around 40,000 observations focusing on the 2016 US political election data.
• Conducted data pre-processing, feature extraction, model building, and evaluation achieving a high accuracy of 98%.
• Built and deployed neural network models.

• Automatic classification of web-scraped clothing data using various neural network methods.
• Utilized 3 RNN techniques to classify over 55,000 observations into 82 classes.
• Achieved accuracy of 87.1% compared to the machine learning baseline model (78.9%).
• Presented findings to stakeholders.

• Updated and communicated effectively with the country manager and CEO through reports, resulting in increased MTN customers.
• Expanded the company’s products through penetrating new markets.
• Oversaw the process of more than 20 products and recommended crucial strategies leading to higher profitability.
• Led the team to achieve highly set targets, resulting in being awarded the employee of th

Data science, machine learning, statistics and operational research.

Earned a Certificate of Higher Education in Computing and Statistics.