Junior AI Engineer, Vision AI

VAELLA; Helsinki; Presencial; EUR 70.000 - 120.000; Jornada completa

Descripción de la vacante

VAELLA is seeking a hands-on ML engineer to improve Vision AI models used in real-world crowded spaces. You will work across data collection, labeling, training and evaluation with on-site work in Helsinki and global travel as needed.

You’ll report to the CTO and collaborate with an international team to build pipelines, generate synthetic data, and deploy models on edge devices in customer environments. English is the working language.

Formación

• Bachelor or master degree in CS/EE/Math or related field.
• Experience training computer vision or deep learning models.
• Good Python skills and ML libraries exposure.
• Ability to work across data collection, labeling, training and deployment.

Responsabilidades

• Collect and label ground truth data on site.
• Generate synthetic training data from simulations.
• Train, evaluate and improve models on edge devices.
• Build and automate data, evaluation and testing pipelines.
• Collaborate with CTO and team on AI coding agents.

Conocimientos

Python

Computer Vision

Deep Learning

English Fluent

Git

Linux

Educación

BSc or MSc in CS/EE/Math

Herramientas

Git

Linux

Docker

Cloud platforms

Descripción del empleo

About VAELLA

An operational brain that turns fragmented data into real-time decisions for the world's busiest spaces.

Why this role is important

Every day, millions of people move through metro stations, airports, ferry terminals, offices and shopping centres, and most of those spaces are still run on assumptions. VAELLA is changing that. Our Vision AI gives crowded spaces a real-time understanding of how people move through them without identifying anyone. We’re building toward a future where infrastructure responds to people as it happens, and where cities become more livable because of it.

Our AI models are at the heart of that future. Every model you train helps our product see more clearly, count more accurately, and work in more places around the world. This is where you build the skills and track record to help shape how cities work.

VAELLA launched in June 2026 as KONE’s first spin-out company. We built our technology from scratch; it’s already running in some of Europe’s most crowded infrastructures. Our ambition is to set the global standard for how crowded spaces understand people, and we’re just getting started.

Our culture

We’re an international team with big ambitions, and we’re building fast. Everyone in our team thinks like an owner: we set big goals, move quickly, and take responsibility for results, not just tasks. You won’t need to wait months to make an impact, because what you build this month can be running in a real building next month.

We share information openly, trust each other to organize our own time, and give people responsibility early. As the company grows, so do the opportunities, and we are looking for people who want to grow with us.

If you’re ambitious, hungry to learn and want to be part of building a company from the ground up, you’ll fit right in.

Your role

You’ll spend most of your time making our Vision AI models better, which is critical to our product’s success. This is a hands-on role across the whole pipeline. One week you might be on site in a real building capturing data, and the next you’re labelling it, running training jobs and checking how the new model performs against ground truth. You’ll report to our CTO, who has 20+years of experience in software and AI, and you’ll learn directly from him every day.

You will

• Collect and label ground truth data on site, the real-world reference our models learn from and are measured against

• Generate synthetic training data from simulated scenes in e.g. NVIDIA Omniverse, so our models learn rare and busy situations before they happen in real life

• Train, evaluate and improve the models that run on edge computers inside our customers’ buildings

• Build and automate the data, evaluation and testing pipelines that tell us whether a model is getting better

• Work with agentic development workflows, using AI coding agents as part of how you build every day

You might be a great fit if

• You have a BSc or MSc in computer science, electrical engineering, mathematics or a related field, or you’re about to graduate

• You have experience training computer vision or deep learning models, through a thesis, project, internship or first job, and you care about how they perform in the real world

• You have good Python skills

• Familiarity with Git and Linux

• You like the whole pipeline: data, labelling, training and testing

• You already use AI coding agents, or you’re eager to work that way

• You're fluent in English, our working language

Nice to have

• Experience with Docker or cloud tools

• Experience with lidar, point clouds, 3D data or simulation

• Willingness to visit customer sites globally

This might not be the best fit if

• You're looking for a management position or want others to do the hands-on, manual work

• You prefer pure research to shipping models that run in real buildings

• You need detailed spec before you start or clearly defined processes to execute – we’re a young company building our own processes as we execute

• You’re looking for a remote position – We're building for the physical world, and we move fastest when we’re together, whether that’s at the whiteboard, in the lab or on site with a customer

Good to know

This role is based in Helsinki. You need a valid right to work in Finland or the EU, as we’re unable to offer visa sponsorship or relocation support.

What to expect during the hiring process

1. Technical interview with our CTO

2. Short assignment/take-home task that we discuss together in a technical interview.

3. Meeting with our whole team

4. Offer

The start date of this role is flexible, and we’re screening applicants on an ongoing basis. The role will be closed as soon as we find the right person. We’ll get back to every applicant by the end of the first week of December 2026.