ML Specialist (Generative AI)
Role in brief
International iGaming Holding seeks an ML Specialist focused on generative AI to develop innovative AI products in a remote setting. Ideal candidates have experience in AI startups and proficiency in Python and PyTorch.
About the role
As an ML Specialist at International iGaming Holding, you will be integral to developing generative AI products, particularly in live communication and content generation. Your role will involve creating generation pipelines for chat messages, images, and videos, while collaborating closely with backend, frontend, product, and design teams.
The company, with over 12 years in the iGaming market, emphasizes a small team environment where decisions are made rapidly and your contributions directly influence product outcomes. You will engage in optimizing inference, fine-tuning models, and conducting experiments, ensuring high-quality results in video generation.
Skills that matter here
- Python: Proficiency in Python is essential for developing AI models and pipelines.
- PyTorch: Practical experience with PyTorch is required for implementing deep learning solutions.
- deep learning: Deep learning expertise is crucial for working with generative models.
- generative models: You will utilize generative models for creating various types of media.
- video generation: Experience in video generation is necessary for the role's focus on content creation.
- fine-tuning: Fine-tuning models will be a key part of optimizing AI outputs.
Who this role suits
- Candidates should have a background in AI startups or products, demonstrating independence in problem-solving.
- A strong understanding of deep learning architectures and video generation techniques is essential.
- The ideal person thrives in a fast-paced environment, enjoys experimenting with new technologies, and can collaborate effectively with cross-functional teams.
From the employer
- Develop generation pipelines using generative AI models: chat messages (LLM), images, and videos
- Work with text-to-video, image-to-video, video-to-video, character-consistent video, lip-sync, motion transfer, face consistency
- Select, test, and fine-tune open-source and proprietary models for product tasks (LoRA, fine-tuning)
- Optimize inference: speed, quality, and cost of generation
- Configure post-processing: upscaling, frame interpolation, face enhancement, moderation, and safety filters
- Conduct experiments and A/B tests for quality
- Participate in building ML infrastructure for video generation and evaluation
- Monitor new models and quickly assess their suitability for the product
- Deliver solutions in production together with backend, frontend, product, and design teams.
- Proficient in Python (not vibe code), practical experience with PyTorch
- 2–3+ years with deep learning and generative models
- Experience in an AI startup or AI product (mandatory)
- Understanding of diffusion and transformer architectures
- Experience with video generation, computer vision, or related fields: lip-sync, identity consistency, motion transfer
- Fine-tuning / LoRA, inference optimization, experience with GPU
- Experience deploying ML in production, understanding the balance of quality / latency / cost
- Independence: propose practical solutions to business problems.
- Remote format, team from different countries
- Salary: Flexible, discussed after the interview
- 28 days of paid vacation, 21 days of paid sick leave, Days Off
- NDA and Job Offer with clear terms
- Real tasks at the forefront of video generation and freedom to try new models.
Questions about this role
Is this position remote?
Yes, this role is fully remote with a team from different countries.
What are the required skills for this role?
Candidates must be proficient in Python and PyTorch, with experience in deep learning and generative models.
How do I apply for this position?
You can apply by visiting the company's website at https://igpgaming.com/.
✓ Apply here in a minute. Your application and answers go straight to the hiring person, so you never land in a spam pile.