Data Analyst
Asan Pardakh
مجموع سنوات الخبرة :4 years, 5 أشهر
As a data analyst at Asan Pardakht, I developed and optimized forecasting services, designed machine learning models, collaborated with teams, and created KPI dashboards for tourism, fintech, and some third-party services reports.
• Designed forecasting services for 6M+ users, utilizing RNNs to achieve over 90% precision, enhancing user experience and decision-making.
• Collaborated with product managers to perform cohort analysis that identified an opportunity to reduce churned users by 20% and boost monthly revenue by $110, 000.
• Led re-engineering of data integration, combining ELT and ETL for 5+ databases, improving efficiency and accuracy, implemented with Apache Airflow for streamlined scheduling.
• Designed user-friendly dashboards in Superset to cover sales, KPIs, user activities, and demographics for a user base of 50 million.
• Conducted A/B testing using Python and statistical methods, optimizing conversion rate by 18%.
• Designed a ML model for calculating Affordability and Credibility for customers. Technology Stack: Python, SQL (PostgreSQL/MySQL), Apache Spark, Kafka, Apache Airflow, Superset, TensorFlow, Keras, Pandas, NumPy, Scikit-learn, Git
As a data analyst at Otaghak, developed recommendation systems and customer behavior models, significantly enhancing user engagement and sales through data-driven strategies.
• Engineered a recommendation system model using Python utilizing tree base model and collaborative filtering/matrix factorization algorithms, leading to a remarkable 65% success ratio increase.
• Deployed machine learning models to predict customer behavior that led to a 19% increase in engagement.
• Conducted customer segmentation through RFM analysis with Python and Pandas, resulting in a substantial 10% sales increase.
• Formulated and executed a Power BI dashboard, enhancing operational efficiency by 25%.
• Analyzed 95, 000 transactional rows using Python to identified 5 distinct customer categories and the formulation of 2 different market strategies. Technology Stack: Python, Scikit-learn, Pandas, NumPy, Power BI, XGBoost, Matplotlib, Seaborn, SQL (MySQL), Git, Jupyter Notebooks
As a junior data scientist at Dayche Group, I worked on image processing and OCR technology projects. I was involved in analyzing medical images and integrating OCR systems to enhance data processing capabilities.
• Analyzed 10, 000 lung images using Python, OpenCV, and Scikit-image to achieve a 95% accuracy in detecting coronavirus and flu infections for the Dycher project.
• Spearheaded the incorporation of an Optical Character Recognition (OCR) system, leveraging Python for data processing, leading to a remarkable 40% processing speed increase. Technology Stack: Python, OpenCV, Scikit-image, TensorFlow, Keras, Tesseract OCR, Pandas, NumPy, Matplotlib, Seaborn, Jupyter Notebooks