Research Engineer / Research Scientist at Ideogram

At Ideogram, we are on a mission to make world-class design accessible to everyone by multiplying human creativity. We specialize in building proprietary generative media models and AI-native creative workflows that solve complex challenges in graphic design. Our team is composed of builders with a proven history of technological breakthroughs, including foundational research in diffusion models and the development of Google’s Imagen and Imagen Video. We value design, taste, and craft just as much as we value rigorous research and engineering. Having raised nearly $100 million from investors like Andreessen Horowitz and Index Ventures, we are scaling rapidly and looking for passionate individuals to help us build the future of the creative economy.

About the Role

We are seeking a Senior Research Engineer or Research Scientist to join our team on a full-time basis. In this role, you will lead the end-to-end development of the post-training pipeline for our text-to-image foundation models. You will guide the strategy from initial data collection through to final deployment, focusing on advancing techniques such as RLHF and RLAIF, as well as working on personalization and customization. You will be an integral part of a research-driven team that values high ownership and collaboration.

Key Responsibilities

• Develop and maintain high-throughput pipelines for fine-tuning and model evaluation.
• Conduct post-training research that drives measurable improvements in model quality and performance.
• Collaborate with a cross-functional team to refine generative AI workflows and push the boundaries of current technology.

Requirements

To be successful in this role, you should have a strong background in deep learning and a passion for generative media. We are looking for the following qualifications:

• At least 5 years of experience developing machine learning models using JAX, PyTorch, or TensorFlow.
• Proven ability to implement machine learning foundations, such as Transformers, VAEs, or Denoising Diffusion models, from scratch.
• A solid track record of innovation in machine learning and a deep understanding of advanced AI techniques.
• Strong debugging skills to iteratively improve model performance and quality.
• Familiarity with Kubernetes and Docker is considered a plus.
• Optional experience in low-level optimization, such as writing CUDA kernel code, is highly valued.

Location

This position is based in Canada.

Compensation and Benefits

We offer a comprehensive package designed to support your well-being and long-term success, including:

• Equity compensation to ensure you share in the growth and impact of our company.
• Medical, dental, and vision insurance coverage starting on your first day.
• 401k retirement savings plan with an employer match.
• Paid time off to ensure you have the space to recharge.