Adobe Firefly's Applied Science & Machine Learning (ASML) group invites an Applied Scientist / Machine Learning Engineer passionate about post-training and distillation of large generative AI models to join the team. This role will focus on raising the quality, efficiency, and deployability of Adobe's generative models for images and videos. The chosen candidate will collaborate with researchers and engineers to build and refine post-training pipelines such as supervised fine-tuning (SFT), preference optimization, and model distillation. These efforts will facilitate adapting large complex models into efficient production versions.
This role influences the performance and scalability of Firefly's generative AI systems, facilitating next-generation creative functionalities for millions of users. As an Applied Scientist at Adobe, you will join a distinguished team of applied researchers and engineers committed to developing and improving generative AI systems. You will work alongside data, modeling, and infrastructure groups to implement post-training upgrades into production systems that support Adobe products.
Job Responsibilities
- Develop and run distillation pipelines to transfer capabilities from large teacher models into smaller, efficient student models.
- Carry out and refine post-training methods including supervised fine-tuning (SFT), preference optimization (DPO/GRPO), and reward-based learning.
- Build infrastructure and tools for teacher rollout creation, distillation data pipelines, and training workflows.
- Carry out experiments aimed at improving model quality, efficiency, and instruction alignment for generative AI models.
- Collaborate closely with research scientists to convert research ideas into scalable training pipelines and production-ready implementations.
- Evaluate models using both automated metrics and human preference signals to guide post-training improvements.
- Optimize models for deployment efficiency, including distillation, model compression, and inference performance.
- Collaborate with various groups such as data, research, and product units to incorporate post-training improvements into Adobe Firefly systems.
What you need to succeed
- Expertise in machine learning algorithms and model distillation techniques
- Strong programming skills in Python or similar languages
- Experience with AI model training and optimization
- Preferred qualifications: Advanced degree or relevant experience in a related field, background in large-scale data pipelines, and knowledge of Adobe products.
Expected Pay Range:Our compensation reflects the cost of labor across several U.S. geographic markets, and we pay differently based on those defined markets. The U.S. pay range for this position is $142,700 -- $270,950 annually. Pay within this range varies by work location and may also depend on job-related knowledge, skills, and experience. Your recruiter can share more about the specific salary range for the job location during the hiring process.
In California, the pay range for this position is $187,100 - $270,950In Washington, the pay range for this position is $168,600 - $244,200
At Adobe, for sales roles starting salaries are expressed as total target compensation (TTC = base + commission), and short-term incentives are in the form of sales commission plans. Non-sales roles starting salaries are expressed as base salary and short-term incentives are in the form of the Annual Incentive Plan (AIP).
In addition, certain roles may be eligible for long-term incentives in the form of a new hire equity award.
