cv engineer

Joi AI is a platform for personalized, emotionally intelligent connections between people and AI characters. The company builds characters with unique personalities, moods, and voices, aiming to facilitate long-term AI-lationships. Its research arm, Joi Lab, develops open-source generative models, agent architectures, and training infrastructure.

• Train and fine-tune generative models (text2image, image2image, video2image, IP-adapters) to produce photorealistic and stylized visuals;
• Work with image classifiers, rankers, and image/video captioning models that understand visual context;
• Track cutting-edge CV research and open-source developments to translate them into a CV roadmap for the ML team;
• Build and own the iterative model improvement process, including quality validation systems for CV outputs;
• Optimize inference performance through diffusion optimization, quantization, and kernel or framework-level work;
• Train multimodal AI companions that combine CV and NLP for deeper visual understanding.

• 5+ Years of experience training diffusion models and modifying their architectures;
• Proficiency with diffusers and transformers libraries;
• Hands-on experience with IP-adapters;
• Backend engineering experience (Python, Go, C#) and knowledge of scalable deployment systems;
• Nice to have: Flow matching training, Diffusion model distillation, DPO and other text2image fine-tuning approaches, Text2video and CLIP fine-tuning, Experience with multimodal LLMs, NLP / LLM training and chatbot development background.

• Standard 28-day annual leave policy;
• 7 Yearly wellness days for life admin or recovery;
• Referral rewards up to $5000;
• 50% Coverage of costs for training, conferences, and global meetups;
• Subsidized English classes;
• Health support up to $1,000 annually for medical fees or private insurance;
• $1000 Every three years for home office or co-working setup;
• Peer-to-peer recognition program with gratitude bonuses.