Senior Machine Learning Engineer

Njuškalo is the largest Croatian marketplace with more than 1.1 million active users. We provide our users a simple and intuitive platform for buying and selling, and our employees with a pleasant business environment, a rich source of knowledge and the opportunity for growth and development. Our culture has been recognized with a Great Place To Work certificate that we obtained for the third year in a row.

We're looking for a Senior ML Engineer who can own machine learning from start to finish, from the first signal all the way to a model running reliably in production. This is a role for someone who is passionate about data and driven by turning messy, large-scale data into models that change how millions of people use a product.

You will be a part of our in-house self-sufficient Data Science team that builds production ML models to make Njuškalo better for every user across our marketplace ecosystem. It's a hands-on role with a lot of ownership, plenty of room to shape the technical direction, and to mentor others along the way.

ABOUT THE ROLE

As our Senior ML Engineer for Multivertical Data Modelling, you will build and own the machine learning models behind our Life Events Framework. The idea is to detect, predict and activate user life-cycle stages across every Njuškalo vertical: Marketplace, Cars & Moto, Real Estate and digital advertising.

Your models turn cross-vertical behavioural signals into something usable: user segments, intent scores and personalisation inputs that feed both the product and the business. You will see it through end to end, from designing the models to running them in production at scale and improving them over time, always within the company's security, quality and GDPR standards.

• →Own the end-to-end development of ML models for the Life Events Framework: signal definition, feature engineering, training, validation and deployment.
• →Build and maintain multivertical user segmentation and intent-prediction models, integrating behavioural data across Marketplace, Cars & Moto, Real Estate and digital advertising.
• →Work with huge volumes of structured and unstructured (image and text) data to deliver the best possible user experience across our platforms.
• →Design detection models, scoring logic and multivariate A/B testing frameworks for life-event use cases.
• →Serve models in a cloud production environment at scale and keep them healthy through monitoring and iteration.
• →Prioritise life-event use cases based on data readiness, revenue potential, cross-vertical impact and speed to MVP.
• →Continuously improve model accuracy through performance tracking, new data integration, user-behaviour insights and additional vertical integrations.
• →Collaborate with vertical sales and product owners, data engineers and the monetisation team to turn life-event signals into product and revenue activation.
• →Research, experiment with and implement state-of-the-art models and techniques.
• →Document the solutions you build, review other ML engineers' code, and mentor junior and mid-level engineers on the team.

Python · PyTorch · MLflow · FastAPI · PostgreSQL · Clickhouse · Elasticsearch & Kibana · Docker & docker-compose · Airflow · dbt · GitLab CI/CD · Linux · AWS · Terraform · Sentry

• 5+ years of machine learning engineering experience.
• Advanced knowledge of machine learning techniques, including deep learning models for computer vision and/or NLP.
• Strong programming skills and advanced command of the frameworks, libraries and databases you work with (PyTorch, TensorFlow, Keras).
• Advanced knowledge of mathematics and statistics.
• Proficiency with SQL for writing efficient queries.
• Hands-on experience building and testing microservices.
• Working knowledge of Linux-based systems and Docker.
• Basic understanding of cloud technologies (AWS).
• Strong problem-solving skills and clear communication.
• A reliable team player who enjoys mentoring and is happy to lead technically.
• Well organised, proactive, and able to manage your own time and priorities.