Machine Learning Engineer

Responsibilities

You will design and operate end-to-end data pipelines and geospatial analytics products using a modern stack including Airflow and Kubernetes. Additionally, you will manage infrastructure and transform analytical outputs into interactive map layers for the platform.

Requirements

The role requires 3+ years of experience in Data Science or Data Engineering with strong proficiency in Python, Docker, and Kubernetes. Candidates must possess deep expertise in SQL, specifically PostGIS, and have a solid understanding of geospatial data and production-grade system ownership.

Skills

• Python
• Airflow
• Kubernetes
• Docker
• PostGIS
• ClickHouse
• Geospatial analytics
• Data pipelines
• ETL
• Mapbox
• SQL
• Infrastructure management
• Visualization engineering
• Spatial statistics

Benefits

• Flexible working hours
• Fully remote role

About this role

RealAdvisor is one of the leading digital platforms in the European real estate ecosystem. Founded in Switzerland, we now support 5,000+ real estate agencies and attract 10+ million users per year across Europe.
We are scaling fast across multiple markets with a clear mission: help real estate professionals grow through data, transparency and smart technology. At RealAdvisor, we value ownership, autonomy and impact.

Role Overview

We are looking for a Machine Learning Engineer to take end-to-end ownership of our geospatial data products.
This is not a model-tuning role. You will design, build, and operate the full data lifecycle behind our real-estate insights — from orchestrating complex pipelines on Kubernetes to delivering high-performance geospatial visualizations with Mapbox.
This is a builder role.
You should be as comfortable debugging an Airflow DAG as you are computing spatial statistics or optimizing analytical queries.
You’ll have full autonomy over a modern, high-performance stack:
Orchestration & Operations: Airflow, Kubernetes, Docker
Data Storage: Postgres (PostGIS) for spatial workloads; ClickHouse for large-scale analytics
Programming: Production-grade Python
Key Responsibilities
End-to-End Pipelines: Design and deploy automated pipelines (ETL) using Airflow to ingest diverse datasets (Real Estate transactions, listings, geo-data).
Geospatial Analytics: Compute advanced spatial metrics (e.g., pricing honeycombs, average prices per neighborhood) using PostGIS and Python.
Infrastructure Management: Maintain and optimize the Kubernetes environment where your pipelines and models run.
Visualization Engineering: Transform your analytical outputs into interactive map layers (Vector Tiles) that integrate seamlessly with Mapbox on our frontend.
What We’re Looking For
Truly Full-Stack (Data): You don’t believe data scientists should stop at notebooks. You own infrastructure, deployment, and reliability because that’s how real value is delivered.
Strong Geospatial Intuition: You understand projections, polygons, spatial joins, and how to visualize density and patterns effectively.
Engineering Mindset: You write clean, modular, maintainable Python code designed for production — not one-off scripts.
Experience: 3+ years in a Data Science or Data Engineering role, with hands-on ownership of production systems.
Infrastructure Proficiency: Solid experience with Docker and Kubernetes, including debugging and troubleshooting live environments.
Database Expertise: Advanced SQL, especially with PostGIS or similar spatial extensions. Experience with ClickHouse is a strong plus.
Orchestration Mastery: Deep hands-on experience with Airflow for building and maintaining reliable data pipelines.
Fully remote role with flexible working hours.
Close collaboration with leadership and go-to-market teams.
International, ambitious, and collaborative team culture.
Fully remote role with flexible working hours.
Close collaboration with leadership and go-to-market teams.
International, ambitious, and collaborative team culture.

About RealAdvisor S.A.

Internet Marketplace Platforms · 51-200 employees · headquartered in Plan-les-Ouates, Geneva · founded 2017