Machine Learning Engineer – Real-Time Video Generation
Full-time | Tech Team
📍 Barcelona (hybrid) or remote within Europe
💰 €45,000 – €55,000
threesixfive is supporting a high-growth conversational AI company building lifelike digital humans for enterprise use cases.
Our client designs and deploys real-time, human-like digital avatars that combine video synthesis, facial animation, voice, and conversational AI. Their technology is used by global organisations to deliver interactive customer experiences at scale, across dozens of languages and markets.
This is a technically ambitious team working at the cutting edge of computer vision, real-time ML systems, and production AI infrastructure, with a strong focus on reliability, performance, and real-world impact.
In a nutshell
Own and scale real-time video synthesis systems for lifelike digital humans
A production-first ML role, bridging research and live deployment
Focus on latency, video quality, and reliability at scale
Why this role stands out
This is a rare opportunity to work on real-time, user-facing ML systems where your optimisations have an immediate and visible impact.
You’ll be part of a small, senior engineering team building cutting-edge digital human technology used by global enterprise customers. The problems are hard, the constraints are real, and the work goes far beyond experimentation or demos.
🚀 The Role
We’re looking for an experienced Production Machine Learning Engineer to take ownership of a real-time video synthesis pipeline.
This is a hands-on, systems-oriented role at the intersection of:
Computer vision
ML infrastructure
Real-time / low-latency systems
You’ll be responsible for taking research models and making them run reliably in production, under strict latency and quality constraints.
🛠️ What You’ll Do
Production Engineering (Core Focus)
Own and operate production video synthesis services
Deploy and optimise ML models for real-time inference
Reduce end-to-end inference latency (targeting sub-2s streaming use cases)
Monitor video quality and system performance; debug production issues
Implement model versioning, A/B testing, and safe rollback strategies
Integration & Optimisation
Act as the bridge between research and production systems
Integrate new CV / video models into an existing real-time pipeline
Design and maintain APIs for video synthesis (gRPC, REST)
Optimise GPU utilisation and inference throughput
Implement caching strategies and support service orchestration
Integrate text-to-speech / audio services where required
Feature Development
Productionise visual improvements (expressiveness, motion, lip-sync)
Support avatar customisation and identity preservation
Turn research prototypes into robust, real-time features
🧰 Tech Stack
Python, PyTorch
AWS, Docker, Kubernetes, GPU-backed instances
gRPC for streaming services
Redis, S3, RabbitMQ
✅ What We’re Looking For
Must-have
3–5 years’ experience deploying ML / CV models into production
Strong hands-on experience with PyTorch or TensorFlow
Practical experience with model optimisation (quantisation, pruning, serving)
Experience with video generation or real-time video processing
Solid understanding of latency vs quality trade-offs
Strong Python skills for backend services and ML serving
Production infrastructure experience (Docker, cloud platforms, CI/CD)
Strong debugging skills and ability to collaborate across teams
Nice-to-have
Experience with audio-driven avatars or face animation
GANs, Diffusion Models, NeRFs, or related CV techniques
gRPC, message queues, or video streaming protocols (HLS, WebRTC)
Open-source contributions or publications in computer vision
🎯 What Success Looks Like
First 6 months
Ownership of core video synthesis services
Improved monitoring, reliability, and operational confidence
At least one research model successfully deployed into production
Measurable improvements in latency or video quality
First 12 months
Significant latency reduction in streaming video delivery (30–50%)
Production rollout of visual improvements
Recognised internally as the go-to engineer for production ML and real-time video systems
💼 What’s on Offer
Compensation & Flexibility
Salary range: €45,000 – €55,000
Hybrid working in Barcelona or remote within Europe
Impact & Growth
End-to-end ownership of critical production systems
Technically challenging problems in real-time ML and scalability
High-impact role in a small, experienced engineering team
Opportunity to influence ML infrastructure as the platform scales
Benefits
Central Barcelona office
Flexible working arrangements
Private health insurance
Commuter / travel allowance
Flexible benefits under local tax schemes
Wellness and fitness discounts
To apply or learn more: email kris@threesixfive.es with your CV or LinkedIn profile.