Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I’m Luka Novak, a Lead Data Scientist based in Zagreb. I lead a data science team of five and design end-to-end ML systems that deliver real business impact, from demand forecasting to customer segmentation and marketing optimization. I enjoy turning data into actionable insights, building scalable pipelines, and mentoring colleagues. My work includes agentic RAG pipelines, Next Best Action platforms, marketing optimization, and customer journey automation, all guided by responsible AI and governance.…Hi, I’m Luka Novak, a Lead Data Scientist based in Zagreb. I lead a data science team of five and design end-to-end ML systems that deliver real business impact, from demand forecasting to customer segmentation and marketing optimization. I enjoy turning data into actionable insights, building scalable pipelines, and mentoring colleagues. My work includes agentic RAG pipelines, Next Best Action platforms, marketing optimization, and customer journey automation, all guided by responsible AI and governance.

Luka Novak

Data Scientist, AI Engineer, Full Stack Developer, +3





Hi, I’m Luka Novak, a Lead Data Scientist based in Zagreb. I lead a data science team of five and design end-to-end ML systems that deliver real business impact, from demand forecasting to customer segmentation and marketing optimization. I enjoy turning data into actionable insights, building scalable pipelines, and mentoring colleagues. My work includes agentic RAG pipelines, Next Best Action platforms, marketing optimization, and customer journey automation, all guided by responsible AI and governance.…Hi, I’m Luka Novak, a Lead Data Scientist based in Zagreb. I lead a data science team of five and design end-to-end ML systems that deliver real business impact, from demand forecasting to customer segmentation and marketing optimization. I enjoy turning data into actionable insights, building scalable pipelines, and mentoring colleagues. My work includes agentic RAG pipelines, Next Best Action platforms, marketing optimization, and customer journey automation, all guided by responsible AI and governance.

Available to hire

Hi, I’m Luka Novak, a Lead Data Scientist based in Zagreb. I lead a data science team of five and design end-to-end ML systems that deliver real business impact, from demand forecasting to customer segmentation and marketing optimization.

I enjoy turning data into actionable insights, building scalable pipelines, and mentoring colleagues. My work includes agentic RAG pipelines, Next Best Action platforms, marketing optimization, and customer journey automation, all guided by responsible AI and governance.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Work Experience

Data Science Team Lead at Cantab PI

September 1, 2021 - Present

Managed data science team of 5. Notable projects: Agentic RAG pipeline for Sales rep assistant using HuggingFace and LangChain; Next Best Action platform driving 30% sales growth in year one; Customer segmentation pipeline with DBSCAN, XGBoost, Optuna, and MLFlow; Marketing channel optimization using GAM reducing spend by 10% and increasing ROI by 5%; Customer journey automation module in Python reducing setup complexity by 80%.

Senior AI/ML Consultant at PricewaterhouseCoopers

August 1, 2020 - September 1, 2021

Led ML projects including OCR-based data extraction with TensorFlow for invoice parsing (audit prep time reduced by 90%); Data quality automation with Python, Pandas, Scikit-learn consolidating 7 dashboards into a single automated report; Supported PwC Global Responsible AI team to define model governance focusing on explainability and bias detection.

AI/ML Consultant at PricewaterhouseCoopers

September 1, 2018 - August 1, 2020

Data reconciliation using DBSCAN and XGBoost; delivered as a web app; Client segmentation and upsell with Scikit-learn and Pandas, generating an additional 4% revenue.

Internships, ML and Data Science at IN2Data; Bejo zaden; IN2; Combis

February 1, 2016 - July 1, 2018

Multiple internships across the Netherlands and Croatia. Projects included tag optimization reducing redundant tags by 40%; article tagging NLP pipeline reducing inconsistencies by 90%; farm management ML-based recommendation engine for precision agriculture.