About the job

Do you want to be part of one of Sweden's fastest-growing start-ups?

Preventai has developed the Antrino platform – a technology that makes every camera intelligent.

Instead of cameras just recording video in the background, they can now perform tasks in real-time: detecting thefts, identifying security risks, analyzing flows in stores and industries, and enabling organizations to act immediately when something happens. Video goes from being something reviewed in retrospect to becoming active, real-time decision support.

This means that millions of existing cameras, which previously only functioned as passive sensors, can suddenly start working actively for businesses – from retail and logistics to security companies and public environments.

Everything is built on a foundation of Privacy by Design, where data protection and GDPR compliance are central parts of the architecture. The goal is to make advanced video analysis both powerful, scalable, and responsible in how data is handled.

Purpose and Goal

We are looking for an AI Engineer who wants to be part of our journey. In this role, you will work on the core intelligence layer of our platform, developing and integrating AI systems that transform video streams into real-time insights and actionable outputs.

Today, parts of our AI stack are rapidly evolving as the platform scales. As we grow, you will play a central role in designing, building, and optimizing AI pipelines, integrating machine learning and video understanding models, and ensuring our systems operate reliably and efficiently in real-world environments.

The role is highly technical and hands-on, where you will work closely with our AI, computer vision, and platform teams to build scalable intelligence systems capable of processing thousands of live video streams in real time. You will help design and improve the architectures, inference pipelines, and data workflows that power our platform, ensuring our AI systems remain robust, efficient, and production-ready as we expand globally.

Your main areas of responsibility will include:

• Designing, developing, and maintaining AI systems and model pipelines that transform video streams into real-time insights
• Building and integrating machine learning and AI models for video understanding, event detection, and intelligent automation
• Developing and optimizing inference pipelines to support low-latency, real-time processing of live video streams
• Improving model performance, robustness, and reliability across diverse real-world environments
• Building and maintaining data pipelines for training, evaluation, and continuous improvement of AI models
• Working with large-scale datasets, including data preparation, labeling strategies, and dataset management
• Optimizing models and pipelines for GPU-accelerated workloads and scalable deployment
• Collaborating closely with computer vision, platform, and DevOps engineers to deploy AI systems in production
• Monitoring model performance in production and iterating based on real-world feedback and system metrics
• Experimenting with and integrating state-of-the-art AI techniques, including multimodal and video-language models
• Contributing to architecture discussions and technical direction of the platform’s AI stack
• Supporting the development of a scalable AI infrastructure capable of processing thousands of live video streams globally

Who are we looking for?

• Strong experience building distributed AI systems, where multiple models operate together in real-time production environments
• Experience orchestrating multi-model inference pipelines using technologies such as Kafka, Triton Inference Server, or similar high-performance model serving systems
• Strong understanding of stream processing, event-driven architectures, and real-time state management
• Experience designing systems where outputs from multiple models (e.g., object detection, tracking, re-identification, and scene understanding) must be correlated and synchronized into unified events
• Ability to build and operate low-latency inference pipelines that handle large-scale video or data streams
• Experience working with GPU-accelerated workloads and optimizing inference performance in production
• Understanding of distributed systems challenges such as throughput, fault tolerance, backpressure, and system observability
• Experience integrating AI systems into scalable production infrastructure rather than isolated research environments
• Strong problem-solving skills and the ability to design robust real-time AI platforms
• Excellent communication skills in English, both written and spoken; Swedish is a plus

Personal qualities

We believe you are a team player with a strong work ethic, who thrives in a fast-paced environment where you build something from the ground up together. You have a positive attitude towards problem-solving and always deliver high quality in your accounting work, while being curious about how processes can be improved and streamlined.

You enjoy solving problems and see opportunities where others see obstacles. In a rapidly growing company, things often change – therefore, you thrive in an environment where initiative, responsibility, and structure are important qualities. You have the ability to work independently, prioritize correctly, and manage several tasks or projects in parallel.

With us, you will be part of a culture where we value intellectual curiosity, high pace, and collaboration between technology and business. We build advanced technology but believe in simple, clear ways of working and a team where everyone contributes to the whole. This means we are looking for people who not only want to do their job – but who also want to help build structures, improve processes, and participate in shaping how the company develops going forward.

Form of employment: Full-time with an initial probationary period of 6 months

Start date: Immediately

Location: Stockholm, Danvikstullen (potential move to a larger office)

Recruitment Process

Our recruitment process is simple and transparent. We review applications continuously and contact candidates we assess to be a good match for the role.

The process usually consists of a first conversation where we get to know each other and go through your background and the role in more detail. This is followed by one or two in-depth conversations where we discuss experience, working methods, and how you could contribute to developing our finance function.

In the final stage, we may also ask for references. Our goal is for the process to be efficient, clear, and respectful of your time, while ensuring that both you and we feel it is the right match.

We look forward to your application!