Looking for a job

Zeynep Acuner

@zeynepacuner

Message

Senior Data Scientist at ONLAYER who built a CatBoost merchant risk-scoring system and production LLM workflows.

Turkey

Message

At ONLAYER, I owned the full machine learning lifecycle for a merchant risk-scoring and portfolio decisioning system, from problem framing through deployment. I built an ensemble of three CatBoost models and a containerised FastAPI inference service, with data-quality checks to catch drift before it reached production.

At UTOPIA MUSIC, I delivered a revenue forecasting model for royalty advances that replaced manual assessment and achieved a 20% uplift in portfolio profitability. I also built audience segmentation models across millions of user records and introduced structured champion/challenger testing.

Through anla, I developed machine learning projects including a multilingual accessibility detector and a verified, multi-stage LLM recommendation pipeline for Hope. Earlier, at NETAŞ, I built predictive maintenance and anomaly detection systems for industrial IoT; my research at Oskar Klein Centre produced four peer-reviewed publications.

Experience

Work history, roles, and key accomplishments

Current

Founder & Developer

Current

Hope

Jan 2025 - Present (1 year 9 months)

Sole technical founder of Hope, a candidate-side job search platform. Built multilingual NLP accessibility detector, regulatory compliance lens, and ghost job detector; developed nightly recommendation pipeline as multi-stage LLM system with verified outputs.

Python React TypeScript fastAPI PostgreSQL Docker GCP Claude APIs LLM NLP EU Regulations

Current

Senior Data Scientist

Current

ONLAYER

Nov 2024 - Present (1 year 11 months)

Owned complete ML lifecycle for merchant risk-scoring and portfolio decisioning system in fintech payment network marketplace. Engineered anomaly detection and schema validation across live data feeds; managed MLflow experiment tracking and built SHAP-based explainability outputs.

Python CatBoost XGBoost scikit learn MLFlow fastAPI Docker PostgreSQL MongoDB SHAP Anomaly Detection GitHub Actions

Freelance Machine Learning Consultant

anla

Mar 2025 - Jan 2026 (10 months)

Delivered consulting projects through anla, a science-driven data analytics and AI studio. Developed NGO misinformation detector, breast cancer prediction model, and scooter demand forecasting.

Machine Learning NLP Classification Demand Forecasting

Data Scientist

UTOPIA MUSIC

Oct 2021 - Oct 2024 (3 years)

Delivered automated revenue forecasting model for royalty advances, achieving 20% uplift in portfolio profitability. Built behavioural audience segmentation models and ran champion/challenger A/B testing programme; managed data infrastructure on GCP.

Python SQL BigQuery GCP Airflow Tableau Docker Forecasting Segmentation A B Testing LLM Claude APIs Hugging Face Transformers AWS Snowflake

Data Scientist

NETAŞ

Nov 2020 - Oct 2021 (11 months)

Built anomaly detection and time-series-based predictive maintenance systems for industrial IoT sensor streams on EUREKA projects. Developed multi-source data collection pipelines and smart-grid load forecasting modules; worked on smart-city applications for Istanbul Municipality.

Python SQL scikit learn Time Series Anomaly Detection Predictive Maintenance IoT Data Pipelines

Researcher

Oskar Klein Centre, KTH

Jan 2015 - Feb 2020 (5 years 1 month)

Led three collaborative research projects applying probabilistic classification and unsupervised clustering on large-scale spectral datasets. Produced four peer-reviewed publications and served as Burst Advocate for the Fermi LAT team.

Python R Statistical Modeling Probabilistic Classification Clustering Hypothesis Testing