About Me
Cem Karisli
MSc Financial Data Analytics | University of Sussex
Education
MSc Financial Data Analytics
University of Sussex, United Kingdom
Focus: Probability, statistics, optimisation, and algorithmic data science with Python-based financial analytics.
MSc Computer and Information Science
Dogus University, Istanbul
Status: Thesis stage
Coursework: Machine Learning, Deep Learning, Computer Vision, Business Intelligence
BSc Industrial Engineering
Cukurova University, Adana
Focus: Operations Research, Optimisation, Statistical Quality Control, Production Planning
Experience
Insurance & Value-Added Services Specialist (Data & Reporting Focus) | MediaMarkt
Reporting automation and performance analytics using Python and advanced Excel.
- Automated recurring business reports with Python and advanced Excel workflows.
- Supported data extraction and analysis to improve reporting speed and consistency.
- Worked in cross-functional agile teams; contributed to delivery tracking and stakeholder updates.
Jan 2025 – Jun 2025
Customer Experience Analyst | Koctas
KPI monitoring and process improvement across contact center and customer experience operations.
- Tracked and reported KPIs such as NPS and operational call metrics; delivered actionable insights.
- Built Excel dashboards and performed root-cause analysis for performance drivers.
- Supported testing and improvement initiatives with internal stakeholders and vendors.
Jan 2023 – Jan 2024
Projects
Cricket Analytics (Industry-Academic Collaboration)
Built baseline expected runs per ball (xRuns) models to support performance evaluation and decision-making.
- Modelled ball-level outcomes using interpretable regression baselines for decision-ready metrics.
- Produced player-level residual leaderboards with uncertainty estimates (standard errors and confidence intervals).
- Delivered report-ready outputs for non-technical decision-makers.
Metric Documentation for ML Classifiers
Interactive Streamlit app visualising and comparing Log Loss and Focal Loss metrics for machine learning classifiers.
- Explained the mathematical intuition and practical use-cases for both metrics.
- Developed interactive charts for exploring metric behaviour across scenarios.
- Deployed on Streamlit Cloud: log-focal-loss.streamlit.app
Pneumonia Detection from X-rays
CNN-based application for pneumonia detection from chest X-ray images, built for a graduate Computer Vision course.
- Implemented a deep learning classifier and evaluated model performance on imbalanced data.
- Created a simple desktop GUI for uploading X-rays and viewing predictions.
Appointment Availability Notifier
Automated monitoring script that checks availability and sends alerts via Telegram to reduce manual effort.
- Built a Python monitoring workflow and notification pipeline.
- Designed it to run on a schedule and send near real-time alerts.
- Focused on reliability and clean logging for maintainability.
Articles
Detecting Credit Card Fraud with Anomaly Detection
Overview of anomaly detection algorithms for identifying fraudulent transactions with machine learning.
Why a Low Pearson Correlation Doesn’t Mean “No Relationship” in Business Data
Exploring why a low Pearson correlation doesn’t always mean variables are unrelated, and how nonlinear business data can mislead analysis.
Fast EDA on Cricket Data with Y-Data Profiling
A fast way to perform Exploratory Data Analysis using Y-Data Profiling in Google Colab, and why interpretation still matters.
Skills
Programming
Core languages and daily tooling.
Data and Machine Learning
Analysis and modelling capabilities.
Visualisation and BI
Communicating insights clearly.
Tools and Platforms
Workflows, collaboration, and data access.