Generative Ai Engineer

Summary

Join our team as a Generative AI Specialist! We are looking for an expert in diffusion-based models (Flux, SD) with hands-on experience in ComfyUI workflow development, model/LORA training, and experimentation. You will develop and optimize workflows, work with various models (Flux, Stable Diffusion), perform LORA training, and train and fine-tune diffusion-based models. Strong Python programming skills and expertise in PyTorch, Diffusers, and Accelerate are essential. Experience deploying AI models and working in a fast-paced environment is required. This is a fully remote position with potential for extra bonuses.

Requirements

• Proficiency with Comfy-UI, Swarm-UI, and model customization
• Deep experience in LORA training and experimentation with Kohya-ss
• Strong Python programming skills
• Ability to design and train required AI models
• Expertise in working with Diffusion models and libraries such as PyTorch, Diffusers, and Accelerate
• Ability to utilize Controlnets, inpainting, Image2Image, to achieve high quality results + other diffusion based techologies/methods
• Strong knowledge of Linux based system
• Proven track record in developing and deploying AI models, training LORAS, designing workflows in comfy
• Willingness to work in a fast-paced environment

Responsibilities

• Develop and optimize workflows using tools like Comfy-UI and Swarm-UI
• Work extensively with models such as Flux dev, Flux schnell, Stable Diffusion (XL, 3, 3.5), and controlnets
• Perform LORA training, focusing on objects, with in-depth experimentation on training parameters
• Train and fine-tune diffusion-based models, leveraging tools (like kohya-ss), python code and libraries like PyTorch, Diffusers, and Accelerate

Preferred Qualifications

• Knowledge of deploying Generative AI solutions
• Experience with Docker and containerization
• Familiarity with cloud platforms like AWS and Google Cloud
• Understanding of Replicate for model deployment

Benefits

• Fully Remote Position
• Extra bonuses may be applicable