Senior Machine Learning Engineer

About ReSpo.Vision

ReSpo.Vision is an AI and Computer Vision company transforming how sports are analyzed, visualized, and monetized. Our proprietary single-camera system extracts elite-level tracking data and performance analytics from standard broadcast or tactical video, without wearables or in-venue installations. Already used by global clubs, federations like FIFA, Concacaf, or CONMEBOL and for competitions like Euro or Copa America, we are actively expanding into media, fan engagement, and betting applications.Our pipeline combines advanced computer vision and deep learning models to track all players and the ball in 3D using a single-camera feed. The resulting positional data powers our growing product suite : from match analytics platform with a LLM layer to visual content, including 3D match reconstruction and real-time broadcast augmentation with dynamic stats and virtual overlays. The system is built for scalability, leveraging cloud-native infrastructure, GPU inference pipelines, and sports-specific post-processing modules that turn raw detections into meaningful insights.

We're now entering an exciting phase focused on two major technical challenges. First, scaling data extraction - we're pushing to process thousands of matches while maintaining reliability and robustness, which means optimizing our pipelines for massive parallel processing and handling diverse conditions from all across the globe Second, we're developing advanced LLM-powered analytics that enable interacting with complex data easily and derivation of insights to help coaches and teams play more effectively. This isn't just about making data accessible - it's about transforming raw positional data into actionable intelligence that wasn't possible before.

Your Role

As a Mid / Senior ML Engineer at ReSpo.Vision , you'll own significant portions of our ML development process, from model architecture through production deployment and system integration. This is a hands-on position where you don't just train models in isolation - you build ML components that seamlessly integrate into our broader sports analytics platform. You'll ship production-ready systems that analyze matches for global sports organizations, always considering how your models interact with upstream data pipelines and downstream applications. You'll take ownership of complete ML workflows, ensuring your solutions scale across hundreds of matches while maintaining the performance standards expected by FIFA, top clubs, and major competitions.

You will be responsible for :

Model Development & System Integration : Design, train, and validate ML models while architecting how they fit into our broader platform

Production-Ready ML : Transform experimental models into robust production systems, handling not just training but also deployment and integration with existing infrastructure

End-to-End Pipeline Ownership : Build complete ML workflows including data preprocessing, model serving, post-processing, and integration with our systems

System-Level Thinking : Consider performance implications, scalability requirements, and downstream dependencies when designing ML solutions - your models need to play nicely with others

Cross-Functional Integration : Collaborate with the rest of the team to ensure smooth model deployment, proper pipeline optimization and assess feature requirements

Production Operations : Own your models in production - monitoring performance, handling edge cases, optimizing inference, and ensuring reliability at scale

Who You Are

You've spent 3-4 years building ML models that actually ship to production and integrate with larger systems

PyTorch is your primary tool and you understand the full journey from research to production deployment

You think beyond model accuracy - considering latency, throughput, resource usage, and system integration

You know the difference between model accuracy on curated test set and within a live production system

Computer vision excites you - detection, segmentation, and working with image / video data at scale

You understand that ML in production is 20% model training and 80% everything else

You can own significant technical areas independently while keeping the big picture in mind

Linux, Docker, and cloud environments (AWS / GCP) are familiar territory

You communicate fluently in English (B2+) and can explain how your ML components fit into complex systems

You are pragmatic - you want to deliver but you know when to stop and think about a problem deeply

You want to push your code to production as quickly as possible

Core Requirements

Strong PyTorch experience with proven ability to take models from prototype to production

Experience with ML lifecycle : data pipelines, training, integration

Algorithmic experience - you don’t run away when you hear BFS, DFS or union-find

Track record of integrating ML models into larger software systems

Solid understanding of MLOps principles and production ML challenges

Experience with computer vision tasks (detection, segmentation) on image and video data

Ability to own end-to-end ML projects while considering system-wide implications

Nice to have

Experience with Hugging Face ecosystem, YOLO, or Detectron2

Background in scalable ML systems and low-latency inference optimization

Knowledge of streaming data systems and real-time ML serving

Comfortable with containerization, cloud platforms, and distributed systems

Experience with model optimization techniques (quantization, pruning, distillation)

Contributions to open-source ML projects

Interest in sports analytics or complex data environments

Experience with distributed training or multi-GPU optimization

What we offer

A chance to work with a top-tier engineering team, including Kaggle Grandmasters

Hybrid work model

Flexibility in employment type (B2B / contract of employment)

Market-competitive salary

Private healthcare and Multisport card

Open training budget – we'll support your development in relevant areas

Ownership and autonomy – no micromanagement, real impact

A unique opportunity to shape a globally recognized, high-impact product used by top sports organizations like Chelsea, Paris Saint-Germain, or FIFA