Resume Details

Self-Employed- Türkiye

2015 – present

AI & Data Scientist| Instructor | Solutions Consultant

Designed, developed, and implemented numerous AI algorithms across various industries; created and delivered a multitude of training content to both local and international customers on-premises and online; analyzed business needs, identified technological solutions, and crafted comprehensive strategic technical visions and plans ready for implementation. Project samples are delivered as follows:

• designed and developed an AI algorithm for ISMEK, an educational institution overseen by the Istanbul Metropolitan Municipality in Türkiye, which achieved a 93% accuracy rate in predicting the likelihood of lifelong learners dropping out of their courses, along with a 31% reduction in learner attrition.
• developed and delivered a specialized training program in AI and Algorithms for HAVELSAN, a company operating in the defense and IT sectors in Türkiye. The program established a solid foundation that resulted in a 24% increase in achieving intelligent outcomes and the production of high-quality products.
• contributed to the design, implementation, and optimization of tailored solutions for the Pakistan Air Force (PAF), aligning them with their unique needs, objectives, and challenges. This effort resulted in a 10% decrease in project delivery delays and a 5% enhancement in the initial project preparation phase.
• developed a specialized algorithm that uses descriptive and predictive modeling to evaluate and predict the daily stock prices of more than 465 companies listed on BIST, the Istanbul Stock Exchange. This algorithm has attained an accuracy rate of about 87% in forecasting daily price fluctuations.
• collaborated with data scientists and engineers from Fauji Fertilizer Company Limited (FCC) to design and develop an AI system for Plant I \& II – Goth Machhi resulting in a 6% reduction in production defects and a 3% increase in overall product quality.
• conducted research and assessment of new machine learning technologies for the R&I division at the National University of Sciences & Technology (NUST), which resulted in the implementation of an advanced deep learning framework, enhancing model performance by 15%.

Probase Road Technology- Kenya

2012 – 2015

BI Developer| International Division

• developed, organized, and maintained operational interfaces for the international division office in Africa, which included dashboards, data visualizations, regular and special reports, and data querying tools. This allowed employees to access essential information from the headquarters in Malaysia and deliver updates on work progress in overseas offices, leading to a 34% increase in office and administration productivity and a 42% reduction in employees misalignment.
• developed and maintained a suite of intelligent software tools for the operational team on the ground, leading to a 15% reduction in implementation defects and a 10% increase in overall project deliveries.
• designed and developed intelligent software tools for analyzing international customer feedback data. This led to the identification of key product improvement areas, resulting in a 15% increase in customer satisfaction, a 23% rise in sales, and a 30% increase in targeted marketing campaign effectiveness.

Limkokwing University- Malaysia

2010 – 2012

Graduate Teaching Assistant

• conducted a variety of teaching and assessment activities, such as tutorials, seminars, and lectures on several occasions.
• contributed to the creation of suitable teaching materials, ensuring that both content and delivery methods aligned with learning objectives, which led to a 27% improvement in student performance over the semester.
• contributed to the supervision of practical work, providing guidance on skills, methods, and techniques that facilitated knowledge transfer, leading to a 12% increase in successful project deliveries.

UNRWA- Jerusalem

2005 – 2010

Microsoft IT Trainer

• conducted a training needs analysis, pinpointing the vital skills needed for different roles among the UNRWA staff in the Jerusalem and Amman offices. This resulted in a 12% enhancement in the on-time performance metric, attributed to the adoption of software tools.
• designed performance assessments to pinpoint skill gaps between the current and desired levels of learner proficiency, resulting in a 30% reduction in skill set misalignment and a 42% enhancement in the outcomes of training programs.
• developed training programs informed by course evaluations, such as eliminating outdated content or introducing new material, which resulted in a 42% enhancement in training effectiveness.

The BireyselValue Algorithm

Link to Project

The developed algorithm offers a novel approach to solving classification problems. It operates on the principle that the distinct characteristics of a class can determine the classification of an observation through similarity metrics. Experimental outcomes from six multiclass datasets across various fields suggest that this method is not only efficient in addressing classification challenges but could also serve as a viable option for feature reduction.

The Big Bang of Data Science

Link to Project

The project has developed and released educational and technical training content centered on five principal domains: research, analysis, prediction, AI programming, and Embedded AI. It includes an overview of quantum mechanics, covering the entire spectrum from concept to execution. With the motto “transitioning from academia to industry,” the project targets a demographic that spans undergraduates, graduates, scholars, researchers, scientists, and IT experts or academicians in computer science or allied fields. The goal is to conceive, construct, and provide solutions that are proficient, practical, and intelligent..

The Norm Culture

Link to Project

This project focuses on the cybersecurity of AI and ML models, specifically introducing two tools: one that simulates two types of training sample poisoning attacks, and another that simulates a defense mechanism against these attacks using a proposed method called ‘norm culture’.

DisguiseData: A Proposed Method for Generating Synthetic Data

Link to Project

Exciting news for data privacy in healthcare, finance, and AI model training! “DisguiseData” is a new method designed to generate synthetic data, mirroring real-world data while safeguarding privacy. This innovative tool empowers users to define parameters like mean variance and deviation levels, creating datasets that closely mirror the original.

2015 – 2019

Ph.D. (Management Information System), KHAS University

GPA: 3.70/4.0

Python listed packages

@dahmansphi