Hi, I'm Onur.
I build data products that forecast, optimize, and drive decisions.
I’m currently leading data science at OODARIS AI, building agentic AI workflows.
Previously: data science at ALDO and OpenSea, research engineering at Nansen, founder at Eigenbeat, and ML foundations at BDC.
Leading data science for an agentic retail platform.
Built and deployed SKU × store allocation optimization with probabilistic forecasting and MIP, improving forecast accuracy and reducing overstock. Designed markdown optimization to uncover multimillion-dollar margin opportunities.
Defined analytics strategy and shipped GCP/dbt pipelines processing 50M+ daily blockchain events to power real-time product metrics and dashboards for 3k+ MAUs.
Owned metrics and dashboards for core marketplace and new products; built wallet labeling and ML heuristics for high-value user tracking to inform growth decisions.
Led DeFi attribution pipelines classifying 30M+ addresses with scalable dbt workflows; deployed ML-based NFT pricing models that improved accuracy for on-chain analytics products.
Built prepayment risk models with LightGBM that drove material savings; created internal Python ML library to standardize modeling workflows across teams.