Microsoft is looking to solve the problem of developing next-gen hardware products that can perceive the user and the user's environment, and is working on several exciting projects that will shape how computers and other devices interact with humans.
Requirements
- 1+ years of experience training/fine tuning AI/ML models, preferably LLMs/SLMs (small language model).
- 1+ years' experience with productization or shipping ML and/or AI components for large-scale internet applications.
- Demonstrated experience with deep learning techniques (e.g., Transformers, RNNs, CNNs, reinforcement learning) and machine learning frameworks (Python, PyTorch, TensorFlow, ONNX).
- Experience working with multimodal data and models (e.g., text, vision, speech, structured signals) and/or retrieval-augmented generation (RAG).
- Proficiency with C/C++ or other performant programming languages a plus.
- Experience creating or deploying evaluation frameworks that drive model improvement, data synthesis, or system-level design.
- Solid background in applied research with ability to translate evaluation insights into new models, end-to-end solutions, and research directions.
Responsibilities
- Design and implement evaluation frameworks to measure the performance of multimodal embeddings, RAG pipelines, and agentic systems.
- Use evaluation insights to drive data synthesis, augmentation, and collection strategies that improve coverage and robustness.
- Build pipelines to test algorithms and models, analyze the results, and translate findings into actionable improvements.
- Develop metrics and benchmarks that inform system-level performance and guide the evolution of end-to-end human-computer interaction solutions.
- Collaborate with researchers and engineers to prototype and scale agentic systems that combine reasoning, planning, and action.
- Research and develop LLMs and SLMs with Python and other relevant programming languages.
- Design and operationalize robust evaluation metrics for multimodal embeddings, retrieval-augmented generation (RAG) pipelines, and agentic systems.
Other
- Bachelor's Degree in Computer Science, Electrical or Computer Engineering, or related field AND 2+ years related experience (e.g., statistics, predictive analytics, research)
- Ability to meet Microsoft, customer and/or government security screening requirements are required for this role.
- Ability to pass the Microsoft Cloud background check upon hire/transfer and every two years thereafter.
- 1+ year(s) experience creating publications (e.g., patents, peer-reviewed academic papers).
- Must be eligible to work in the United States without sponsorship.
