Curriculum Vitae

Data Scientist and Business Growth Professional with over 5 years of experience driving measurable outcomes across multiple industries through advanced analytics, statistical modeling, and machine learning. Goes beyond analysis to translate data into strategic decisions that directly impact revenue, operational efficiency, and growth. Delivered results including a 32% increase in production volume, a 72% accurate demand forecasting model, and a 15% reduction in product shelf time. Combines deep quantitative expertise with a strong commercial mindset to create impact at the intersection of data and business strategy.

• 08/2024 — Present: Data Science & Business Growth Analyst @ Dedeoglu Jewelry & Goldsmith Co., Turkey
  • Developed an AI model that predicts the timing and magnitude of customer demand shifts driven by gold price volatility with 72% accuracy; model outputs were directly integrated into investment gold inventory management and procurement decisions.
  • Combined data from major e-commerce marketplaces, competitor analysis, Google Trends, and offline customer demand to conduct product-level demand analysis; the analysis was directly utilized by management to reduce average product shelf time by 15%.
  • Contributed to the design of advertising campaigns that increased the conversion rate from 0.74% to 0.96%; collaborated with the e-commerce and marketing teams to segment the online customer base and develop segment-specific campaign strategies.
  • Designed and implemented a survey measuring visit frequency, purchasing habits, spending levels, and satisfaction; combined survey data with RFM segmentation to classify customer profiles and contributed to the development of segment-specific customer retention strategies.
  • Analyzed financial and operational datasets covering sales performance, inventory costs, and marketing expenditures; developed interactive dashboards in Google Looker Studio and communicated findings to business partners through regular and ad-hoc presentations. These efforts directly supported data-driven decision-making in advertising budget optimization, inventory cost reduction, and sales performance improvement.
• 01/2024 — 06/2024: AI/ML Engineer (Contract) @ Miuul, Sweden (Remote)
  • Mined historical WhatsApp and Discord conversations to extract student questions and produced domain-specific answers, building a natural training dataset from scratch.
  • Diagnosed insufficient model performance and designed a synthetic data generation pipeline using GPT-4o; restructured training by combining natural and synthetic datasets.
  • This initiative drove an 18% improvement in model accuracy, measured via LLM-as-a-Judge evaluation (1–5 scoring scale).
  • Applied QLoRA fine-tuning to open-source Llama-2 variants, experimenting across quantization levels to identify the optimal configuration.
  • Played a key role at the intersection of model development and AI engineering within a 7-person cross-functional team.
• 06/2019 — 11/2023: Quantitative Genetics & Data Analyst @ Kaymaz Livestock Production, Turkey
  • Led end-to-end data-driven projects in a livestock breeding operation, applying statistical modeling and experimental design to drive measurable business outcomes.
  • Collected and processed livestock performance data through systematic field observations and measurements, managing datasets across multiple generations using spreadsheets.
  • Built a quantitative genetics pipeline using BLUPF90 — merged multi-generational pedigree records to expand a constrained dataset, estimated breeding values (EBV), and selected genetically superior animals — resulting in a 32% increase in total annual production volume.
  • Designed and executed hypothesis-driven experiments (t-test, ANOVA, REML, MCMC, GLM) using SAS, Excel, and Python to validate breeding decisions and ensure continuous product improvement.
  • Diversified revenue streams from 2 to 3 channels (live animal sales, beef, raw milk) by leveraging genetic improvement to increase live weight and introducing raw milk as a recurring income source — growing total revenue by ~20% within 4.5 years.
  • Reduced feed costs by 27% through data-informed pasture development and optimized feed formulations.

• Cat-Dog Classifier App — Computer Vision
  Developed a computer vision classification app using pre-trained architectures (MobileNet, VGG16, ResNet50) with TensorFlow and Keras. Tracked experiments with MLflow and deployed a prototype interface via Streamlit. Achieved 97.14% classification accuracy.
• Apache vs The Fraudster — Big Data & Fraud Detection
  Built a fraud detection pipeline on a distributed file system using Apache Hadoop and Hive on Linux/Unix. Developed and optimized a machine learning model on Apache Spark with Git for version control. Achieved 78.8% recall and 90.0% precision.
• Cite Me if You Can — Generative AI & Semantic Search
  Built a semantic search and summarization tool for scientific documents using RAG and SPECTER2 embeddings. Developed a FastAPI backend integrated with Gemini 2.5 Flash via Google Generative AI API for source-grounded summaries. Designed a Streamlit UI supporting file upload, similarity search, and auto-generated citations in APA/MLA/Chicago formats.

• 02/2020 — 08/2022: Master of Science @ Çanakkale Onsekiz Mart University, Graduate School of Natural and Applied Sciences
  Focused on statistics, data analytics, and quantitative genetics. Led experimental research on survivability traits across 64,000 animals using advanced statistical methods (regression, cluster analysis, hypothesis testing, forecasting).
  • Thesis: Factors Affecting Survivability in Japanese Quails (Coturnix coturnix japonica)
• 08/2015 — 01/2020: Bachelor of Science @ Çanakkale Onsekiz Mart University, Faculty of Agricultural Engineering, Department of Animal Science
  Major in Animal Science, minor in Animal Nutrition.

• Statistics & Quantitative Analysis
  Statistical modeling, hypothesis testing (t-test, ANOVA, GLM), regression analysis, experimental design, REML, MCMC, time-series analysis, exploratory data analysis (EDA)
• Machine Learning & Predictive Modeling
  Supervised/unsupervised learning, LightGBM, Random Forest, Linear/Logistic Regression, model validation and optimization, predictive analytics
• Deep Learning & Computer Vision
  TensorFlow, Keras, ANN, CNN, RNN, LSTM, transfer learning (MobileNet, VGG16, ResNet50)
• Programming & Data Manipulation
  Python (NumPy, Pandas, Scikit-Learn, Matplotlib, Seaborn, Statsmodels, SciPy), SQL, data cleaning & preprocessing
• Data Visualization & Business Intelligence
  Google Looker Studio, Power BI, dashboard development, stakeholder reporting
• AI & NLP
  Large language models (LLMs), QLoRA fine-tuning, RAG, NLP, transformers (BERT, GPT), LangChain, LangGraph
• Data Engineering & Tools
  ETL pipelines, web scraping (Selenium, BeautifulSoup), Docker, Git, MLflow, FastAPI
• Cloud & Big Data
  Microsoft Azure, Apache Hadoop, Hive, Apache Spark, HiveQL

• 08/2023 — 12/2023: Data Scientist Bootcamp @ Miuul
  Intensive bootcamp covering machine learning, data analysis, and end-to-end pipeline development in Python.
• 03/2024: AI Engineering Professional Certificate @ IBM
  Covers Machine Learning with Python, Deep Learning with Keras, Computer Vision, Deep Neural Networks with PyTorch, Building Deep Learning Models with TensorFlow, and AI Capstone Project.

• Estimation of (Co)Variance Components for Viability, Incubation Length and Hatch Weight in Japanese Quail – 13th National Animal Science Conference, 2023, Ankara.
• The Effect of Poultry Red Mite (Dermanyssus gallinae) Infestation on the Viability of Japanese Quail – 13th National Animal Science Conference, 2023, Ankara.
• Factors Affecting Survivability in Japanese Quails (Coturnix coturnix japonica), 2022.
• Plumage color, hatching day and hatching weight influence the survivability up to sexual maturity in Japanese Quail – 3rd International and 12th National Animal Science Conference, 2021, Bursa.

• 12/2023 — Present: Co-Founder & Community Lead @ Data Jungle
  • Established and currently co-leading a community of dedicated data science enthusiasts committed to continuous improvement in the fields of data science and artificial intelligence.
  • Developing applications powered by machine learning and/or generative artificial intelligence.
  • Conducting research and experiments on machine learning algorithms.
  • Delivering presentations on data science projects, new technologies, and statistical methods.
  • Facilitating effective communication between team members, coordinating meetings, assigning tasks, and fostering an environment of open communication to lead collaborative efforts towards shared objectives.