top of page

AI Engineer (Data Scientist / Computer Vision)

Ljubljana, Slovenia

Job Type

Full Time

About the Role

We are looking for a computer vision engineer, who is not afraid to tackle various types of challenges, implement the latest state-of-the-art and come up with new ideas.

Your main responsibilities will be the development of new convolutional neural networks for semantic segmentation and object detection on 3D point clouds and images, training them and running them in production.


  • implement the latest state-of-the-art,

  • train, validate and optimise AI Models,

  • come up with a novel solution to improve the performance of existing AI models,

  • write efficient data ingestion pipelines,

  • write technical documentation.


  • experience with implementing AI Models in deep learning frameworks such as Tensorflow or PyTorch,

  • experience with development in Python,

  • not afraid of 3D data,

  • basic Linux (ssh into a remote host, etc.),

  • experience using GIT,

  • experience with Docker,

  • ability to learn quickly,

  • ability to write clear and concise code that your teammates can understand and build upon,

  • a drive to understand what is going on,

  • comfortable with professional software engineering best practices.

Nice to have

  • experience with 3D data and point clouds,

  • experience with AWS,

  • drive to tackle geospatial problems that can help make this world a better place.

What we offer

  • full-time employment,

  • stimulative compensations,

  • modern offices in Ljubljana,

  • remote work and flexible hours,

  • work with highly motivated and skilled engineers,

  • choose your preferred hardware (Mac/PC),

  • challenging and rewarding tasks,

  • work on a product that helps better understand and manage our planet,

  • personal and career development.

About the Company

Flai is a spinoff company of Flycom Technologies.

Flai develops a web-based application that automatically turns raw geospatial data into meaningful information. Point clouds, images, and satellite data are processed via advanced AI models – this complex work is done automatically on a scalable infrastructure which reduces time-to-process from months to days.

Our AI models are pretrained on multiple data sources and can be used for processing right away. Also, our solution enables additional training of AI models. All annotations done on the web application are automatically used to generate better results. Retraining the AI model greatly reduces the time and cost needed for quality control and manual edits.

To apply send us an email at 

bottom of page