Overview
We’re hiring a Senior Applied AI Engineer, Image Generation to join a fast‑moving, high‑ownership team building next‑generation AI assistant and productivity capabilities. This role blends LLM product engineering, evaluation science, hillclimbing, and internal tool building with the pace and creativity of a startup.
The primary focus for this role will be on building the best image generation product out there. You'll get to experiment with all sorts of models from open source to SOTA developed at Microsoft. You'll work through some of the most challenging multimodal problems that exist today while shipping improvements to customers daily.
Responsibilities
Model Development & Training
- Train, fine-tune, and evaluate image generation models (diffusion, GAN, transformer-based)
- Implement and adapt techniques from research papers into working production systems
- Design and run experiments to improve image quality, diversity, and controllability
- Curate, clean, and manage large-scale image-text training datasets
Evaluation, Hillclimbing & Quality Systems
Build and maintain evaluation frameworks for correctness, safety, grounding, and UX quality.
- Run hillclimbing loops across prompts, models, and tool‑use strategies to continuously improve assistant performance.
- Analyze failure modes, design mitigations, and drive systematic improvements across the stack.
LLM Tooling & Internal Infrastructure
- Develop internal tools for prompt experimentation, model comparison telemetry and debugging automated eval pipelines
- Create reusable frameworks that accelerate the entire AI org’s ability to ship high‑quality assistant features.
Applied ML & Product Integration
- Integrate LLMs with product surfaces, APIs, and backend systems.
- Build lightweight ML components (ranking, classification, summarization, personalization) that enhance assistant intelligence.
- Collaborate with PM, design, and research to turn ambiguous ideas into polished user experiences.
High‑Velocity Teamwork
Operate with startup‑founder energy: bias for action, rapid iteration, and comfort with ambiguity.
- Work closely with researchers, engineers, and product leaders in a fast‑moving AI team where ideas ship quickly and impact is immediate.
- Contribute to a culture of experimentation, clarity, and high‑quality execution.
- Build prompt architectures, system instructions, and orchestration logic that ensure reliability, grounding, and personality consistency.
- Prototype new capabilities rapidly and iterate based on user signals and evaluation data.
Production & Infrastructure
- Optimize models for inference latency, throughput, and cost (quantization, distillation, batching)
- Build and maintain serving pipelines for real-time and batch image generation
- Develop APIs and SDKs that expose image generation capabilities to downstream teams/products
- Monitor model performance in production; debug quality regressions
Qualifications
Required Qualifications
- Bachelor's Degree in Computer Science or related technical field AND 4+ years technical engineering experience with coding in languages including, but not limited to, C, C++, C#, Java, JavaScript, or Python OR equivalent experience.
Preferred Qualifications
- Master’s Degree AND 3+ years of experience in engineering, problem solving, model building, evaluation, data analysis OR equivalent experience.
- PhD in engineering, applied math, statistics, or related analytical field.
- 2+ years shipping production-level code, models, or data analysis.
- 1+ years using AI-assisted coding and analysis techniques.
- Solid grasp of deep learning: loss functions, optimization, regularization, training stability
- Experience deploying ML models at scale (inference optimization, quantization, distillation)
- Familiarity with image preprocessing pipelines, data augmentation, and dataset curation
- Experience working on small teams and mid-stage startup environments.
- Experience working on AI products.
- 4+ years shipping production-level code, models, or data analysis.
- Deep experience building from zero-to-one.
- Experience with RLHF / DPO for aligning image models to human preferences
- Knowledge of safety/content filtering for generated images
- Hands on work hillclimbing AI evaluations.
Data Science IC4 - The typical base pay range for this role across the U.S. is USD $119,800 - $234,700 per year. There is a different range applicable to specific work locations, within the San Francisco Bay area and New York City metropolitan area, and the base pay range for this role in those locations is USD $158,400 - $258,000 per year.
Certain roles may be eligible for benefits and other compensation. Find additional benefits and pay information here: https://careers.microsoft.com/us/en/us-corporate-pay
Data Science IC5 - The typical base pay range for this role across the U.S. is USD $139,900 - $274,800 per year. There is a different range applicable to specific work locations, within the San Francisco Bay area and New York City metropolitan area, and the base pay range for this role in those locations is USD $188,000 - $304,200 per year.
Certain roles may be eligible for benefits and other compensation. Find additional benefits and pay information here: https://careers.microsoft.com/us/en/us-corporate-pay
This position will be open for a minimum of 5 days, with applications accepted on an ongoing basis until the position is filled.
Microsoft is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to age, ancestry, citizenship, color, family or medical care leave, gender identity or expression, genetic information, immigration status, marital status, medical condition, national origin, physical or mental disability, political affiliation, protected veteran or military status, race, ethnicity, religion, sex (including pregnancy), sexual orientation, or any other characteristic protected by applicable local laws, regulations and ordinances. If you need assistance with religious accommodations and/or a reasonable accommodation due to a disability during the application process, read more about **requesting accommodations.**