Arda KAYMAZ

I'm a Data Scientist and Business Growth Analyst with over 5 years of hands-on experience building data solutions that directly drive revenue, efficiency, and strategic decisions.

My career spans three distinct industries, which gives me both depth and range. At Dedeoglu Jewelry, I built an AI model to predict customer demand shifts tied to gold price volatility (72% accuracy), integrated directly into procurement decisions. I also led product-level demand analysis across major e-commerce platforms, helping reduce average shelf time by 15%, and designed customer segmentation strategies that improved ad conversion from 0.74% to 0.96%.

Before that, I worked as an AI/ML Engineer at Miuul (Sweden, remote), where I developed a domain-specific LLM-powered chatbot for technical student support. I designed a synthetic data generation pipeline using GPT-4o and applied QLoRA fine-tuning to Llama-2 variants, achieving an 18% improvement in model accuracy.

Earlier in my career, I led end-to-end data projects in livestock breeding at a family agribusiness, applying quantitative genetics, BLUPF90 pipelines, and statistical modeling (REML, MCMC, ANOVA) to achieve a 32% increase in annual production volume and ~20% revenue growth.

I hold an M.Sc. in Animal Science (Statistics, Data Analytics & Quantitative Genetics) and have published research on survivability modeling across 64,000 animal species. I'm also a certified IBM AI Engineering Professional and completed an intensive Data Science Bootcamp at Miuul.

Technically, I work across the full ML stack: Python, SQL, TensorFlow, LightGBM, LangChain, RAG, FastAPI, Docker, Apache Spark, and Power BI/Looker Studio. I'm equally comfortable in model development, AI engineering, and stakeholder communication.

I'm currently looking for remote or hybrid opportunities in data science, ML engineering, or applied AI — ideally at a company where data has real impact on business outcomes.